Download and Read online, DOWNLOAD EBOOK, [PDF EBOOK EPUB ], Ebooks

download, Read Ebook EPUB/KINDE, Download Book Format PDF

Bioinformatics Research and Applications 12th

International Symposium ISBRA 2016 Minsk Belarus
June 5 8 2016 Proceedings 1st Edition Anu
Bourgeois

OR CLICK LINK
https://textbookfull.com/product/bioinformatics-
research-and-applications-12th-international-
symposium-isbra-2016-minsk-belarus-
june-5-8-2016-proceedings-1st-edition-anu-
bourgeois/
Read with Our Free App Audiobook Free Format PFD EBook, Ebooks dowload PDF
with Andible trial, Real book, online, KINDLE , Download[PDF] and Read and Read
Read book Format PDF Ebook, Dowload online, Read book Format PDF Ebook,
[PDF] and Real ONLINE Dowload [PDF] and Real ONLINE
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Bioinformatics Research and Applications 10th

International Symposium ISBRA 2014 Zhangjiajie China
June 28 30 2014 Proceedings 1st Edition Mitra Basu

https://textbookfull.com/product/bioinformatics-research-and-
applications-10th-international-symposium-isbra-2014-zhangjiajie-
china-june-28-30-2014-proceedings-1st-edition-mitra-basu/

Experimental Algorithms 15th International Symposium

SEA 2016 St Petersburg Russia June 5 8 2016 Proceedings
1st Edition Andrew V. Goldberg

https://textbookfull.com/product/experimental-algorithms-15th-
international-symposium-sea-2016-st-petersburg-russia-
june-5-8-2016-proceedings-1st-edition-andrew-v-goldberg/

Pattern Recognition and Information Processing: 13th

International Conference, PRIP 2016, Minsk, Belarus,
October 3-5, 2016, Revised Selected Papers 1st Edition
Viktor V. Krasnoproshin
https://textbookfull.com/product/pattern-recognition-and-
information-processing-13th-international-conference-
prip-2016-minsk-belarus-october-3-5-2016-revised-selected-
papers-1st-edition-viktor-v-krasnoproshin/

Theory and Applications of Satisfiability Testing SAT

2016 19th International Conference Bordeaux France July
5 8 2016 Proceedings 1st Edition Nadia Creignou

https://textbookfull.com/product/theory-and-applications-of-
satisfiability-testing-sat-2016-19th-international-conference-
bordeaux-france-july-5-8-2016-proceedings-1st-edition-nadia-
Applied Reconfigurable Computing 12th International
Symposium ARC 2016 Mangaratiba RJ Brazil March 22 24
2016 Proceedings 1st Edition Vanderlei Bonato

https://textbookfull.com/product/applied-reconfigurable-
computing-12th-international-symposium-arc-2016-mangaratiba-rj-
brazil-march-22-24-2016-proceedings-1st-edition-vanderlei-bonato/

Engineering Secure Software and Systems 8th

International Symposium ESSoS 2016 London UK April 6 8
2016 Proceedings 1st Edition Juan Caballero

https://textbookfull.com/product/engineering-secure-software-and-
systems-8th-international-symposium-essos-2016-london-uk-
april-6-8-2016-proceedings-1st-edition-juan-caballero/

NASA Formal Methods 8th International Symposium NFM

2016 Minneapolis MN USA June 7 9 2016 Proceedings 1st
Edition Sanjai Rayadurgam

https://textbookfull.com/product/nasa-formal-methods-8th-
international-symposium-nfm-2016-minneapolis-mn-usa-
june-7-9-2016-proceedings-1st-edition-sanjai-rayadurgam/

Unifying Theories of Programming 6th International

Symposium UTP 2016 Reykjavik Iceland June 4 5 2016
Revised Selected Papers 1st Edition Jonathan P. Bowen

https://textbookfull.com/product/unifying-theories-of-
programming-6th-international-symposium-utp-2016-reykjavik-
iceland-june-4-5-2016-revised-selected-papers-1st-edition-
jonathan-p-bowen/

Search Based Software Engineering 8th International

Symposium SSBSE 2016 Raleigh NC USA October 8 10 2016
Proceedings 1st Edition Federica Sarro

https://textbookfull.com/product/search-based-software-
engineering-8th-international-symposium-ssbse-2016-raleigh-nc-
usa-october-8-10-2016-proceedings-1st-edition-federica-sarro/
 Anu Bourgeois · Pavel Skums
Xiang Wan · Alex Zelikovsky (Eds.)
LNBI 9683

Bioinformatics Research
and Applications
12th International Symposium, ISBRA 2016
Minsk, Belarus, June 5–8, 2016
Proceedings

123
Lecture Notes in Bioinformatics 9683

Subseries of Lecture Notes in Computer Science

LNBI Series Editors

Sorin Istrail
Brown University, Providence, RI, USA
Pavel Pevzner
University of California, San Diego, CA, USA
Michael Waterman
University of Southern California, Los Angeles, CA, USA

LNBI Editorial Board

Søren Brunak
Technical University of Denmark, Kongens Lyngby, Denmark
Mikhail S. Gelfand
IITP, Research and Training Center on Bioinformatics, Moscow, Russia
Thomas Lengauer
Max Planck Institute for Informatics, Saarbrücken, Germany
Satoru Miyano
University of Tokyo, Tokyo, Japan
Eugene Myers
Max Planck Institute of Molecular Cell Biology and Genetics, Dresden,
Germany
Marie-France Sagot
Université Lyon 1, Villeurbanne, France
David Sankoff
University of Ottawa, Ottawa, Canada
Ron Shamir
Tel Aviv University, Ramat Aviv, Tel Aviv, Israel
Terry Speed
Walter and Eliza Hall Institute of Medical Research, Melbourne, VIC, Australia
Martin Vingron
Max Planck Institute for Molecular Genetics, Berlin, Germany
W. Eric Wong
University of Texas at Dallas, Richardson, TX, USA
More information about this series at http://www.springer.com/series/5381
Anu Bourgeois Pavel Skums
•

Xiang Wan Alex Zelikovsky (Eds.)

•

Bioinformatics Research
and Applications
12th International Symposium, ISBRA 2016
Minsk, Belarus, June 5–8, 2016
Proceedings

123
Editors
Anu Bourgeois Xiang Wan
Department of Computer Science Department of Computer Science
Georgia State University Hong Kong Baptist University
Atlanta, GA Kowloon Tong
USA Hong Kong SAR
Pavel Skums Alex Zelikovsky
Centre for Disease Control and Prevention Georgia State University
Atlanta, GA Atlanta, GA
USA USA

ISSN 0302-9743 ISSN 1611-3349 (electronic)

Lecture Notes in Bioinformatics
ISBN 978-3-319-38781-9 ISBN 978-3-319-38782-6 (eBook)
DOI 10.1007/978-3-319-38782-6

Library of Congress Control Number: 2016938416

LNCS Sublibrary: SL8 – Bioinformatics

© Springer International Publishing Switzerland 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are
believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors
give a warranty, express or implied, with respect to the material contained herein or for any errors or
omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
 Preface

The 12th edition of the International Symposium on Bioinformatics Research and

Applications (ISBRA 2016) was held during June 5–8, 2016, in Minsk, Belarus. The
symposium provides a forum for the exchange of ideas and results among researchers,
developers, and practitioners working on all aspects of bioinformatics and computa-
tional biology and their applications.
There were 77 submissions received in response to the call for papers. The Program
Committee decided to accept 42 of them for publication in the proceedings and for oral
presentation at the symposium: 22 for Track 1 (an extended abstract) and 20 for Track
2 (an abridged abstract). The technical program also featured invited keynote talks by
five distinguished speakers: Dr. Teresa M. Przytycka from the National Institutes of
Health discussed the network perspective on genetic variations, from model organisms
to diseases; Prof. Ion Mandoiu from the University of Connecticut spoke on challenges
and opportunities in single-cell genomics; Prof. Alexander Schoenhuth from Centrum
Wiskunde and Informatica spoke on dealing with uncertainties in big genome data;
Prof. Ilya Vakser from the University of Kansas discussed genome-wide structural
modeling of protein–protein interactions; and Prof. Max Alekseyev from George
Washington University spoke on multi-genome scaffold co-assembly based on the
analysis of gene orders and genomic repeats.
We would like to thank the Program Committee members and external reviewers for
volunteering their time to review and discuss symposium papers. Furthermore, we
would like to extend special thanks to the steering and general chairs of the symposium
for their leadership, and to the finance, publicity, local organization, and publication
chairs for their hard work in making ISBRA 2016 a successful event. Last but not least,
we would like to thank all authors for presenting their work at the symposium.

June 2016 Anu Bourgeois

Pavel Skums
Xiang Wan
Alex Zelikovsky
 Organization

Steering Chairs
Dan Gusfield University of California, Davis, USA
Ion Măndoiu University of Connecticut, USA
Yi Pan Georgia State University, USA
Marie-France Sagot Inria, France
Ying Xu University of Georgia, USA
Alexander Zelikovsky Georgia State University, USA

General Chairs
Sergei V. Ablameyko Belarusian State University, Belarus
Bernard Moret École Polytechnique Fédérale de Lausanne,
Switzerland
Alexander V. Tuzikov National Academy of Sciences of Belarus

Program Chairs
Anu Bourgeois Georgia State University, USA
Pavel Skums Centers for Disease Control and Prevention, USA
Xiang Wan Hong Kong Baptist University, SAR China
Alex Zelikovsky Georgia State University, USA

Finance Chairs
Yury Metelsky Belarusian State University, Belarus
Alex Zelikovsky Georgia State University, USA

Publications Chair
Igor E. Tom National Academy of Sciences of Belarus

Local Arrangements Chair

Dmitry G. Medvedev Belarusian State University, Belarus

Publicity Chair
Erin-Elizabeth A. Durham Georgia State University, USA
VIII Organization

Workshop Chair
Ion Mandoiu University of Connecticut, USA

Program Committee
Kamal Al Nasr Tennessee State University, USA
Sahar Al Seesi University of Connecticut, USA
Max Alekseyev George Washington University, USA
Mukul S. Bansal University of Connecticut, USA
Robert Beiko Dalhousie University, Canada
Paola Bonizzoni Università di Milano-Bicocca, Italy
Anu Bourgeois Georgia State University, USA
Zhipeng Cai Georgia State University, USA
D.S. Campo Rendon Centers for Disease Control and Prevention, USA
Doina Caragea Kansas State University, USA
Tien-Hao Chang National Cheng Kung University, Taiwan
Ovidiu Daescu University of Texas at Dallas, USA
Amitava Datta University of Western Australia, Australia
Jorge Duitama International Center for Tropical Agriculture, Columbia
Oliver Eulenstein Iowa State University, USA
Lin Gao Xidian University, China
Lesley Greene Old Dominion University, USA
Katia Guimaraes Universidade Federal de Pernambuco, Brazil
Jiong Guo Universität des Saarlandes, Germany
Jieyue He Southeast University, China
Jing He Old Dominion University, USA
Zengyou He Hong Kong University of Science and Technology,
SAR China
Steffen Heber North Carolina State University, USA
Jinling Huang East Carolina University, USA
Ming-Yang Kao Northwestern University, USA
Yury Khudyakov Centers for Disease Control and Prevention, USA
Wooyoung Kim University of Washington Bothell, USA
Danny Krizanc Wesleyan University, USA
Yaohang Li Old Dominion University, USA
Min Li Central South University, USA
Jing Li Case Western Reserve University, USA
Shuai Cheng Li City University of Hong Kong, SAR China
Ion Mandoiu University of Connecticut, USA
Fenglou Mao University of Georgia, USA
Osamu Maruyama Kyushu University, Japan
Giri Narasimhan Florida International University, USA
Bogdan Pasaniuc University of California at Los Angeles, USA
Steven Pascal Old Dominion University, USA
Andrei Paun University of Bucharest, Romania
 Organization IX

Nadia Pisanti Università di Pisa, Italy

Russell Schwartz Carnegie Mellon University, USA
Joao Setubal University of Sao Paulo, Brazil
Xinghua Shi University of North Carolina at Charlotte, USA
Yi Shi Jiao Tong University, China
Pavel Skums Centers for Disease Control and Prevention, USA
Ileana Streinu Smith College, USA
Zhengchang Su University of North Carolina at Charlotte, USA
Wing-Kin Sung National University of Singapore, Singapore
Sing-Hoi Sze Texas A&M University, USA
Weitian Tong Georgia Southern University, USA
Gabriel Valiente Technical University of Catalonia, Spain
Xiang Wan Hong Kong Baptist University, SAR China
Jianxin Wang Central South University, China
Li-San Wang University of Pennsylvania, USA
Peng Wang Shanghai Advanced Research Institute, China
Lusheng Wang City University of Hong Kong, SAR China
Seth Weinberg Old Dominion University, USA
Fangxiang Wu University of Saskatchewan, Canada
Yufeng Wu University of Connecticut, USA
Minzhu Xie Hunan Normal University, China
Dechang Xu Harbin Institute of Technology, China
Can Yang Hong Kong Baptist University, SAR China
Ashraf Yaseen Texas A&M University-Kingsville, USA
Guoxian Yu Southwest University, China
Alex Zelikovsky Georgia State University, USA
Yanqing Zhang Georgia State University, USA
Fa Zhang Institute of Computing Technology, China
Leming Zhou University of Pittsburgh, USA
Fengfeng Zhou Chinese Academy of Sciences, China
Quan Zou Xiamen University, China

Additional Reviewers

Aganezov, Sergey Cirjila, Ionela Leung, Yuk Yee Previtali, Marco

Alexeev, Nikita Cobeli, Matei Li, Xin Rizzi, Raffaella
Avdeyev, Pavel Della Vedova, Luo, Junwei Sun, Sunah
Castro-Nallar, Gianluca Pei, Jingwen Sun, Luni
Eduardo Ghinea, Razvan Peng, Xiaoqing Wang, Weixin
Chen, Huiyuan Ionescu, Vlad Pham, Son Zhang, June
Chen, Yong Kuksa, Pavel Postavaru, Stefan Zhao, Wei
Chu, Chong Lara, James Preda, Catalina Zhou, Chan
 Contents

Next Generation Sequencing Data Analysis

An Efficient Algorithm for Finding All Pairs k-Mismatch Maximal

Common Substrings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Sharma V. Thankachan, Sriram P. Chockalingam, and Srinivas Aluru

Poisson-Markov Mixture Model and Parallel Algorithm for Binning

Massive and Heterogenous DNA Sequencing Reads . . . . . . . . . . . . . . . . . . 15
Lu Wang, Dongxiao Zhu, Yan Li, and Ming Dong

FSG: Fast String Graph Construction for De Novo Assembly of Reads Data . . . 27
Paola Bonizzoni, Gianluca Della Vedova, Yuri Pirola, Marco Previtali,
and Raffaella Rizzi

OVarCall: Bayesian Mutation Calling Method Utilizing Overlapping

Paired-End Reads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Takuya Moriyama, Yuichi Shiraishi, Kenichi Chiba, Rui Yamaguchi,
Seiya Imoto, and Satoru Miyano

High-Performance Sensing of DNA Hybridization on Surface

of Self-organized MWCNT-Arrays Decorated by Organometallic
Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
V.P. Egorova, H.V. Grushevskaya, N.G. Krylova, I.V. Lipnevich,
T.I. Orekhovskaja, B.G. Shulitski, and V.I. Krot

Towards a More Accurate Error Model for BioNano Optical Maps . . . . . . . . 67

Menglu Li, Angel C.Y. Mak, Ernest T. Lam, Pui-Yan Kwok, Ming Xiao,
Kevin Y. Yip, Ting-Fung Chan, and Siu-Ming Yiu

HapIso: An Accurate Method for the Haplotype-Specific Isoforms

Reconstruction from Long Single-Molecule Reads. . . . . . . . . . . . . . . . . . . . 80
Serghei Mangul, Harry (Taegyun) Yang, Farhad Hormozdiari,
Elizabeth Tseng, Alex Zelikovsky, and Eleazar Eskin

Protein-Protein Interactions and Networks

Genome-Wide Structural Modeling of Protein-Protein Interactions. . . . . . . . . 95

Ivan Anishchenko, Varsha Badal, Taras Dauzhenka, Madhurima Das,
Alexander V. Tuzikov, Petras J. Kundrotas, and Ilya A. Vakser

Identifying Essential Proteins by Purifying Protein Interaction Networks . . . . 106

Min Li, Xiaopei Chen, Peng Ni, Jianxin Wang, and Yi Pan
XII Contents

Differential Functional Analysis and Change Motifs in Gene Networks

to Explore the Role of Anti-sense Transcription . . . . . . . . . . . . . . . . . . . . . 117
Marc Legeay, Béatrice Duval, and Jean-Pierre Renou

Predicting MicroRNA-Disease Associations by Random Walking on

Multiple Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Wei Peng, Wei Lan, Zeng Yu, Jianxin Wang, and Yi Pan

Progression Reconstruction from Unsynchronized Biological Data using

Cluster Spanning Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Ryan Eshleman and Rahul Singh

Protein and RNA Structure

Consistent Visualization of Multiple Rigid Domain Decompositions

of Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Emily Flynn and Ileana Streinu

A Multiagent Ab Initio Protein Structure Prediction Tool for Novices

and Experts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Thiago Lipinski-Paes, Michele dos Santos da Silva Tanus,
José Fernando Ruggiero Bachega, and Osmar Norberto de Souza

Filling a Protein Scaffold with a Reference. . . . . . . . . . . . . . . . . . . . . . . . . 175

Letu Qingge, Xiaowen Liu, Farong Zhong, and Binhai Zhu

Phylogenetics

Mean Values of Gene Duplication and Loss Cost Functions . . . . . . . . . . . . . 189

Paweł Górecki, Jarosław Paszek, and Agnieszka Mykowiecka

The SCJ Small Parsimony Problem for Weighted Gene Adjacencies . . . . . . . 200
Nina Luhmann, Annelyse Thévenin, Aïda Ouangraoua, Roland Wittler,
and Cedric Chauve

Path-Difference Median Trees. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

Alexey Markin and Oliver Eulenstein

NEMo: An Evolutionary Model with Modularity for PPI Networks . . . . . . . . 224

Min Ye, Gabriela C. Racz, Qijia Jiang, Xiuwei Zhang,
and Bernard M.E. Moret

Multi-genome Scaffold Co-assembly Based on the Analysis of Gene Orders

and Genomic Repeats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Sergey Aganezov and Max A. Alekseyev
 Contents XIII

Sequence and Image Analysis

Selectoscope: A Modern Web-App for Positive Selection Analysis

of Genomic Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Andrey V. Zaika, Iakov I. Davydov, and Mikhail S. Gelfand

Methods for Genome-Wide Analysis of MDR and XDR Tuberculosis from

Belarus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Roman Sergeev, Ivan Kavaliou, Andrei Gabrielian, Alex Rosenthal,
and Alexander Tuzikov

Haplotype Inference for Pedigrees with Few Recombinations . . . . . . . . . . . . 269

B. Kirkpatrick

Improved Detection of 2D Gel Electrophoresis Spots by Using Gaussian

Mixture Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
Michal Marczyk

Abridged Track 2 Abstracts

Predicting Combinative Drug Pairs via Integrating Heterogeneous Features

for Both Known and New Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
Jia-Xin Li, Jian-Yu Shi, Ke Gao, Peng Lei, and Siu-Ming Yiu

SkipCPP-Pred: Promising Prediction Method for Cell-Penetrating Peptides

Using Adaptive k-Skip-n-Gram Features on a High-Quality Dataset . . . . . . . 299
Wei Leyi and Zou Quan

CPredictor2.0: Effectively Detecting both Small and Large Complexes from

Protein-Protein Interaction Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Bin Xu, Jihong Guan, Yang Wang, and Shuigeng Zhou

Structural Insights into Antiapoptotic Activation of Bcl-2 and Bcl-xL

Mediated by FKBP38 and tBid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
Valery Veresov and Alexander Davidovskii

VAliBS: A Visual Aligner for Bisulfite Sequences . . . . . . . . . . . . . . . . . . . 307

Min Li, Xiaodong Yan, Lingchen Zhao, Jianxin Wang, Fang-Xiang Wu,
and Yi Pan

MegaGTA: A Sensitive and Accurate Metagenomic Gene-Targeted

Assembler Using Iterative de Bruijn Graphs . . . . . . . . . . . . . . . . . . . . . . . . 309
Dinghua Li, Yukun Huang, Henry Chi-Ming Leung, Ruibang Luo,
Hing-Fung Ting, and Tak-Wah Lam
XIV Contents

EnhancerDBN: An Enhancer Prediction Method Based on Deep Belief

Network. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Hongda Bu, Yanglan Gan, Yang Wang, Jihong Guan,
and Shuigeng Zhou

An Improved Burden-Test Pipeline for Cancer Sequencing Data . . . . . . . . . . 314

Yu Geng, Zhongmeng Zhao, Xuanping Zhang, Wenke Wang, Xiao Xiao,
and Jiayin Wang

Modeling and Simulation of Specific Production of Trans10,

cis12-Conjugated Linoleic Acid in the Biosynthetic Pathway . . . . . . . . . . . . 316
Aiju Hou, Xiangmiao Zeng, Zhipeng Cai, and Dechang Xu

Dynamic Protein Complex Identification in Uncertain Protein-Protein

Interaction Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
Yijia Zhang, Hongfei Lin, Zhihao Yang, and Jian Wang

Predicting lncRNA-Protein Interactions Based on Protein-Protein Similarity

Network Fusion (Extended Abstract) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
Xiaoxiang Zheng, Kai Tian, Yang Wang, Jihong Guan,
and Shuigeng Zhou

DCJ-RNA: Double Cut and Join for RNA Secondary Structures

Using a Component-Based Representation . . . . . . . . . . . . . . . . . . . . . . . . . 323
Ghada Badr and Haifa Al-Aqel

Improve Short Read Homology Search Using Paired-End Read Information . . . 326
Prapaporn Techa-Angkoon, Yanni Sun, and Jikai Lei

Framework for Integration of Genome and Exome Data for More Accurate
Identification of Somatic Variants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
Vinaya Vijayan, Siu-Ming Yiu, and Liqing Zhang

Semantic Biclustering: A New Way to Analyze and Interpret Gene

Expression Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
Jiří Kléma, František Malinka, and Filip Železný

Analyzing microRNA Epistasis in Colon Cancer Using Empirical Bayesian

Elastic Nets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
Jia Wen, Andrew Quitadamo, Benika Hall, and Xinghua Shi

Tractable Kinetics of RNA–Ligand Interaction . . . . . . . . . . . . . . . . . . . . . . 337

Felix Kühnl, Peter F. Stadler, and Sebastian Will

MitoDel: A Method to Detect and Quantify Mitochondrial DNA Deletions

from Next-Generation Sequence Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
Colleen M. Bosworth, Sneha Grandhi, Meetha P. Gould,
and Thomas LaFramboise
 Contents XV

TRANScendence: Transposable Elements Database and De-novo Mining

Tool Allows Inferring TEs Activity Chronology . . . . . . . . . . . . . . . . . . . . . 342
Michał Startek, Jakub Nogły, Dariusz Grzebelus, and Anna Gambin

Phylogeny Reconstruction from Whole-Genome Data Using Variable

Length Binary Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
Lingxi Zhou, Yu Lin, Bing Feng, Jieyi Zhao, and Jijun Tang

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347

Next Generation Sequencing Data
Analysis
 An Efficient Algorithm for Finding All Pairs
k-Mismatch Maximal Common Substrings

Sharma V. Thankachan1 , Sriram P. Chockalingam2 , and Srinivas Aluru1(B)

1
School of CSE, Georgia Institute of Technology, Atlanta, USA
sharma.thankachan@gatech.edu, aluru@cc.gatech.edu
2
Department of CSE, Indian Institute of Technology, Bombay, India
sriram.pc@iitb.ac.in

Abstract. Identifying long pairwise maximal common substrings

among a large set of sequences is a frequently used construct in com-
putational biology, with applications in DNA sequence clustering and
assembly. Due to errors made by sequencers, algorithms that can accom-
modate a small number of differences are of particular interest, but
obtaining provably efficient solutions for such problems has been elusive.
In this paper, we present a provably efficient algorithm with an expected
run time guarantee of O(N logk N + occ), where occ is the output size,
for the following problem: Given a collection D = {S1 , S2 , . . . , Sn } of n
sequences of total length N , a length threshold φ and a mismatch thresh-
old k ≥ 0, report all k-mismatch maximal common substrings of length
at least φ over all pairs of sequences in D. In addition, we present a result
showing the hardness of this problem.

1 Introduction

Due to preponderance of DNA and RNA sequences that can be modeled as

strings, string matching algorithms have myriad applications in computational
biology. Modern sequencing instruments sequence a large collection of short reads
that are randomly drawn from one or multiple genomes. Deciphering pairwise
relationships between the reads is often the first step in many applications. For
example, one may be interested in finding all pairs of reads that have a sufficiently
long overlap, such as suffix/prefix overlap (for genomic or metagenomic assem-
bly) or substring overlap (for read compression, finding RNA sequences contain-
ing common exons, etc.). Sequencing instruments make errors, which translate to
insertion, deletion, or substitution errors in the reads they characterize, depend-
ing on the type of instrument. Much of modern-day high-throughput sequencing
is carried out using Illumina sequencers, which have a small error rate (< 1–2 %)
and predominantly (> 99 %) substitution errors. Thus, algorithms that tolerate
a small number of mismatch errors can yield the same solution as the much
more expensive alignment/edit distance computations. Motivated by such appli-
cations, we formulate the following all pairs k -mismatch maximal common
substrings problem:


c Springer International Publishing Switzerland 2016
A. Bourgeois et al. (Eds.): ISBRA 2016, LNBI 9683, pp. 3–14, 2016.
DOI: 10.1007/978-3-319-38782-6 1
4 S.V. Thankachan et al.

Problem 1. Given a collection D = {S1 , S2 , . . . , Sn } of n sequences of total

length N , a length threshold φ, and a mismatch threshold k ≥ 0, report all
k-mismatch maximal common substrings of length ≥ φ between any pair of
sequences in D.

Throughout this paper, for a sequence Si ∈ D, |Si | is its length, Si [x] is its
xth character and Si [x..y] is its substring starting at position x and ending at
position y, where 1 ≤ x ≤ y ≤ |Si |. For brevity, we may use Si [x..] to denote the
suffix of Si starting at x. The substrings Si [x..(x+l −1)] and Sj [y..(y +l −1)] are
a k-mismatch common substring if the Hamming distance between them is at
most k. It is maximal if it cannot be extended on either side without introducing
another mismatch.
Before investigating Problem 1, consider an existential version of this prob-
lem, where we are just interested in enumerating those pairs (Si , Sj ) that contain
at least one k-mismatch maximal common substring of length ≥ φ. For k = 0 and
any given (Si , Sj ) pair, this can be easily answered in O(|Si | + |Sj |) time using a
generalized suffix tree based algorithm because no mismatches are involved. For
k ≥ 1, we can use the recent result by Aluru et al. [1], by which the k-mismatch
longest common substring can be identified in O((|Si | + |Sj |) logk (|Si | + |Sj |))
time. Therefore, this straightforward approach  can solve the existential prob-
lem over all pairs of sequences in D in i j (|Si | + |Sj |) logk (|Si | + |Sj |) =
O(n2 L logk L) time, where L = maxi |Si |. An interesting question is, can we
solve this problem asymptotically faster?
We present a simple result showing that the existence of a purely combi-
natorial algorithm, which is “significantly” better than the above approach is
highly unlikely in the general setting. In other words, we prove a conditional
lower bound, showing that the bound is tight within poly-logarithmic factors.
This essentially shows the hardness of Problem 1, as it is at least as hard as its
existential version. Our result is based on a simple reduction from the boolean
matrix multiplication (BMM) problem. This hard instance is simulated via care-
ful tuning of the parameters – specifically, when L approaches the number of
sequences and φ = Θ(log n). However, this hardness result does not contradict
the possibility of an O(N + occ) run-time algorithm for Problem 1, because occ
can be as large as Θ(n2 L2 ).
In order to solve such problems in practice, the following seed-and-extend type
filtering approaches are often employed (see [13] for an example). The underline
principle is: if two sequences have a k-mismatch common substring of length ≥ φ,
then they must have an exact common substring of length at least τ =  k+1 φ
.
Therefore, using some fast hashing technique, all pairs of sequences that have a
τ -length common substring are identified. Then, by exhaustively checking all such
candidate pairs, the final output is generated. Clearly, such an algorithm cannot
provide any run time guarantees and often times the candidate pairs generated
can be overwhelmingly larger than the final output size. In contrast to this, we
develop an O(N ) space and O(N logk N +occ) expected run time algorithm, where
occ is the output size. Additionally, we present the results of some preliminary
experiments, in order to demonstrate the effectiveness of our algorithm.
 An Efficient Algorithm for Finding All Pairs 5

2 Notation and Preliminaries

Let Σ be the alphabet for all sequences in D. Throughout the paper, both |Σ| and
k are assumed to be constants. Let T = S1 $1 S2 $2 . . . Sn $n be the concatenation
of all sequences in D, separated by special characters $1 , $2 , . . . , $n . Here each
$i is a unique special symbol and is lexicographically larger than all characters
in Σ. Clearly, there exists a one to one mapping between the positions in T
(except the $i positions) and the positions in the sequences in D. We use lcp(·, ·)
to denote the longest common prefix of two input strings and lcpk (·, ·) to denote
their longest common prefix while permitting at most k mismatches. We now
briefly review some standard data structures that will be used in our algorithms.

2.1 Suffix Trees, Suffix Arrays and LCP Data Structures

The generalized suffix tree of D (equivalently, the suffix tree of T), denoted by
GST, is a lexicographic arrangement of all suffixes of T as a compact trie [11,15].
The GST consists of |T| leaves, and at most (|T| − 1) internal nodes all of which
have at least two child nodes each. The edges are labeled with substrings of T.
Let path of u refer to the concatenation of edge labels on the path from root
to node u, denoted by path(u). If u is a leaf node, then its path corresponds
to a unique suffix of T (equivalently a unique suffix of a unique sequence in D)
and vice versa. For any node, node-depth is the number of its ancestors and
string-depth is the length of its path.
The suffix array [10], SA, is such that SA[i] is the starting position of the suffix
corresponding to the ith left most leaf in the suffix tree of T, i.e., the starting
position of the ith lexicographically smallest suffix of T. The inverse suffix array
ISA is such that ISA[j] = i, if SA[i] = j. The Longest Common Prefix array LCP
is defined as, for 1 ≤ i < |T|

LCP[i] = |lcp(TSA[i] , TSA[i+1] )|

In other words, LCP[i] is the string depth of the lowest common ancestor of
the ith and (i + 1)th leaves in the suffix tree. There exist optimal sequential
algorithms for constructing all these data structures in O(|T|) space and time
[6,8,9,14].
All operations on GST required for our purpose can be simulated using
SA, ISA, LCP array, and a range minimum query (RMQ) data structure over
the LCP array [2]. A node u in GST can be uniquely represented by an interval
[sp(u), ep(u)], the range corresponding to the leaves in its subtree. The string
depth of u is the minimum value in LCP[sp(u), ep(u) − 1] (can be computed
in constant time using an RMQ). Similarly, the longest common prefix of any
two suffixes can also be computed in O(1) time. Finally, the k-mismatch longest
common prefix of any two suffixes can be computed in O(k) time as follows: let
l = |lcp(T[x..], T[y..])|, then for any k ≥ 1, |lcpk (T[x..], T[y..])| is also l if either of
T[x+l], T[y+l] is a $i symbol, else it is l+1+|lcpk−1 (T[(x+l+1)..], T[(y+l+1)..])|.
6 S.V. Thankachan et al.

2.2 Linear Time Sorting of Integers and Strings

Our algorithm relies heavily on sorting of integers and strings (specifically, suf-
fixes). The integers are always within the range [1, N ]. Therefore, linear time
sorting (via radix sort) is possible when the number of integers to be sorted is
sufficiently large (say ≥ N  for some constant  > 0). In order to achieve linear
sorting complexity for even smaller input sizes, we use the following strategy:
combine multiple small collections such that the total size becomes sufficiently
large. Assign a unique integer id in [1, N ] to each small collection, and with each
element e within a small collection with id i, associate the key i · N + e. We can
now apply radix sort and sort all elements within the combined collection w.r.t.
their associated key. An additional scanning step suffices to separate all small
collections with their elements sorted. The strings to be sorted in our algorithms
are always suffixes of T. Using GST, each suffix can be mapped to an integer
in [1, N ] that corresponds to its lexicographic rank in constant time. Therefore,
any further suffix sorting (or sparse suffix sorting) can also be reduced to linear
time integer sorting.

3 Hardness Result
3.1 Boolean Matrix Multiplication (BMM)
The input to BMM problem consists of two n × n boolean matrices A and B
with all entries either 0 or 1. The task is to compute their boolean product C.
Specifically, let Ai,j denotes the entry corresponding to ith row and jth column
of matrix A, where 0 ≤ i, j < n, then for all i, j, compute


n−1
Ci,j = (Ai,k ∧ Bk,j ). (1)
k=0

Here ∨ represents the logical OR and ∧ represents the logical AND opera-
tions. The time complexity of the fastest matrix multiplication algorithm is
O(n2.372873 ) [16]. The result is obtained via algebraic techniques like Strassen’s
matrix multiplication algorithm. However, the fastest combinatorial algorithm is
better than the standard cubic algorithm by only a poly-logarithmic factor [17].
As matrix multiplication is a highly studied problem over decades, any truly
sub-cubic algorithm for matrix multiplication which is purely combinatorial is
highly unlikely, and will be a breakthrough result.

3.2 Reduction
In this section, we prove the following result.

Theorem 1. If there exists an f (n, L) time algorithm for the existential version
of Problem 1, then there exists an O(f (2n, O(n log n))) + O(n2 ) time algorithm
for the BMM problem.
 An Efficient Algorithm for Finding All Pairs 7

This implies BMM can be solved in sub-cubic time via purely combinatorial
techniques if f (n, L) is O(n2− L) or O(n2 L1− ) for any constant  > 0. As the
former is less likely, the later is also less likely. In other words, O(n2 L) bound is
tight under BMM hardness assumption. We now proceed to show the reduction.
We create n sequences X1 , X2 , . . . , Xn from A and n sequences Y1 , Y2 , . . . , Yn
from B. Strings are of length at most L = nlog n + n − 1 over an alphabet
Σ = {0, 1, $, #}. The construction is the following.

– For 1 ≤ i ≤ n, let Ui = {k | Ai,k = 1}. Then, for each Ui , create a correspond-

ing string Xi as follows: encode each element in Ui in binary in log n bits,
append it with $ symbol, then concatenate all of them. For example, if n = 4
and Ui = {0, 2}, then Xi = 00$10$.
– For 1 ≤ i ≤ n, let Vj = {k | Bk,j = 1}. Then, for each Vj , create a corre-
sponding string Yj as follows: encode each element in Vj in binary in log n
bits, append it with # symbol, then concatenate all of them. For example, if
n = 4 and Vj = {1, 2}, then Yj = 01#10#.
– The database D is the set of all Xi ’s and Yj ’s.

Lemma 1. The entry Ci,j = 1, iff Ui ∩ Vj is not empty. The set Ui ∩ Vj is not

empty iff Xi and Yj have a common substring of length log n.
Using this result, we can compute the boolean product of A and B from the out-
put of the existential version of Problem 1 plus additional O(n2 ) time. Therefore,
BMM can be solved in f (2n, O(n log n)) + O(n2 ) time.

4 Our Algorithm for k-Mismatch Maximal Common

Substrings
We present an O(N logk N + occ) expected time algorithm. Recall that each
position x in the concatenated text T corresponds to a unique position in a
unique sequence Sd ∈ D, therefore we denote the sequence identifier d by seq(x).
Each output can be represented as a pair (x, y) of positions in T, where
1. seq(x) = seq(y)
2. T[x − 1] = T[y − 1]
3. |lcpk (T[x..], T[y..])| ≥ φ.

4.1 The Exact Match Case

Without mismatches, the problem can be easily solved in optimal O(N + occ)
worst case time. First create the GST, then identify all nodes whose string depth
is at least φ, such that the string depth of their parent nodes is at most (φ−1). Such
nodes are termed as marked nodes. Clearly, a pair of suffixes satisfies condition (3)
iff their corresponding leaves are under the same marked node. This allows us to
process the suffixes under each marked node w independently as follows: let Suff w
denotes the set of starting positions of the suffixes of T corresponding to the leaves
in the subtree of w. That is, Suff w = {SA[j] | sp(w) ≤ j ≤ ep(w)}. Then,
8 S.V. Thankachan et al.

1. Partition Suff w into (at most) Σ + 1 buckets, such that for each σ ∈ Σ,
there is a unique bucket containing all suffixes with previous character σ.
All suffixes with their previous character is a $ symbol are put together in a
special bucket. Note that a pair (x, y) is an answer only if both x are y are
not in the same bucket, or if both of them are in the special bucket.
2. Sort all suffixes w.r.t the identifier of the corresponding sequence (i.e., seq(·)).
Therefore, within each bucket, all suffixes from the same sequence appear
contiguously.
3. For each x, report all answers of the from (x, ·) as follows: scan every bucket,
except the bucket in which x belongs to, unless x is in the special bucket.
Then output (x, y) as an answer, where y is not an entry in the contiguous
chunk of all suffixes from seq(x).

The construction of GST and Step (1) takes O(N ) time. Step (2) over all
marked nodes can also be implemented in O(N ) time via integer sorting (refer
to Sect. 2.2). By noting down the sizes of each chunk during Step (2), we can
implement Step (3) in time proportional to the sizes of input and output. By
combining all, the total time complexity is O(N · |Σ| + occ), i.e., O(N + occ)
under constant alphabet size assumption.

4.2 The k-Mismatch Case

Approximate matching problems are generally harder, because the standard data
structures such as suffix trees and suffix arrays may not apply directly. There-
fore, we follow a novel approach, where we transform the approximate matching
problem over exact strings into an exact matching problem over inexact copies
of strings, which we call as modified suffixes.

Definition 1. Let # be a special symbol not in Σ. A k-modified suffix is a suffix

of T with its k characters replaced by #.

Let Δ be a set of positions. Then, TΔ [x..] denotes the |Δ|-modified suffix

obtained by replacing |Δ| positions in the suffix T[x..] as specified by Δ. For
example, let T = aaccgattcaa, Δ = {2, 4}, then T[5..] = gattcaa and TΔ [5..] =
g#t#caa.
Our algorithm consists of two main phases. In the first phase, we create a
collection of sets of k-modified suffixes. In the second phase, we independently
process each set constructed in the first phase and extract the answers. The
first phase takes O(N H k ) time, where as the second phase takes O(N H k + occ)
time. Here H is the height of GST. It is known that the expected value of H is
O(log N ) [3]. Therefore, by combining the time complexities of both phases with
H replaced by O(log N ), we obtain the expected run time as claimed. We now
describe these phases in detail.

Details of Phase-1. This phase is recursive (with levels of recursion starting

from 0 up to k), such that at each level h > 0, we create a collection of sets of
 An Efficient Algorithm for Finding All Pairs 9

h-modified suffixes (denoted by C1h , C2h , . . . ) from the sets in the previous level.
At level 0, we have only a single set C01 , the set of all suffixes of T. See Fig. 1 for
an illustration. To generate the sets at level h, we take each set Cgh−1 at level
(h − 1) and do the following:

– Create a compact trie of all strings in Cgh−1

– For each internal node w in the trie, create a set consisting of the strings
corresponding to the leaves in the subtree of w, but with their (l + 1)th
character replaced by #. Here l is the string depth of w. Those strings with
their (l + 1)-th character is a $i symbol are not included.

level 0 C10

level 1 C11 C21 . . . . . .

.
.
.
level (h − 1)
. . h−1
Cg−1 Cgh−1 h−1
Cg+1 . .

level (h)
. . Cfh−1 Cfh Cfh+1 . .

level k C1k C2k C3k C4k . . . .

Fig. 1. The sets . . . , Cfh−1 , Cfh , Cfh+1 . . . of h-modified suffixes are generated from the
set Cgh−1 .

From our construction procedure, the following properties can be easily

verified.

Property 1. All modified suffixes within the same set (at any level) have #
symbols at the same positions and share a common prefix at least until the last
occurrence of #.

Property 2. For any pair (x, y) of positions, there will be exactly one set
at level k, such that it contain k-modified suffixes of T[x..] and T[y..] with #
symbols at the first k positions in which they differ. Therefore, the lcp of those
k-modified suffixes is equal to the lcpk of T[x..] and T[y..].

We have the following result about the sizes of these sets.

Lemma 2. No set is of size more than N and the sum of sizes of all sets at a
particular level h is ≤ N × H k , where H is the height of GST.
10 S.V. Thankachan et al.

Proof. The first statement follows easily from our construction procedure and the
second statement can be proved via induction. Let Sh be the sum of sizes of all
sets at level h. Clearly, the base case, S0 = |C10 | = N , is true. The sum of sizes of
sets at level h generated from Cgh−1 is at most |Cgh−1 |× the height of the compact
trie over the strings in Cgh−1 . The height of the compact trie is ≤ H, because if we
remove the common prefix of all strings in Cgh−1 , they are essentially suffixes of
T. By putting these together, we have Sh ≤ Sh−1 · H ≤ Sh−2 · H 2 ≤ · · · ≤ N H k .
Space and Time Analysis: We now show that Phase-1 can be implemented in
O(N ) space and O(N logk N ) time in expectation. Consider the step where we
generate sets from Cgh−1 . The lexicographic ordering between any two (h − 1)-
modified suffixes in Cgh−1 can be determined in constant time. i.e., by simply
checking the lexicographic ordering between those suffixes obtained by deleting
their common prefix up to the last occurrence of #. Therefore, suffixes can
be sorted using any comparison based sorting algorithm. After sorting, the lcp
between two successive strings can be computed in constant time. Using the
sorted order and lcp values, we can construct the compact trie using standard
techniques in the construction of suffix tree [4]. Finally, the new sets can be
generated in time proportional to their size. In summary, the time complexity
for a particular Cgh−1 is O(|Cgh−1 |(log N + H)). Overall time complexity is
 
Cfk + (H + log n) Cfh = O(N (log N + H)k−1 H)
h≤k f h<k f

By replacing H by O(log N ), we bound the expect run time of Phase-1 by

O(N logk N ).
We generate the sets in pre-order of the corresponding node in the recursion
tree. As soon as a set (at level k) is generated, we immediately pass it to Phase-2,
extract the necessary information and discard it from the working space. Also,
any set at level h < k is also deleted after all k-level sets in its subtree are
processed. This way, at any point of time in the execution of the algorithm, we
need to maintain only k sets, corresponding to the sets in a root to leaf path in
the recursion tree. Since the size of each set is at most N (Lemma 2), we can
bound the working space also by O(N ), assuming k = O(1).

Details of Phase-2. In this phase, we seek to process each set Cfk created by
Phase-1 independently and generate the answers in time linear to the total size of
all sets and output. i.e., O(N H k +occ). We first present a simple O((N +occ)H k )
time approach. Following are the key steps.

1. Create a compact trie over all k-modified suffixes in Cfk . Then identify the
marked nodes as before. Recall that a marked node has a string depth ≥ φ,
where as the string depth of its parent is < φ.
2. Let Δ be the set of k positions corresponding to modifications in the k-
modified suffixes in Cfk . Clearly, if the leaves corresponding to two modified
suffixes (say TΔ [x..] and TΔ [y..]) are in the same subtree of a marked node,
 An Efficient Algorithm for Finding All Pairs 11

then their lcpk is ≥ φ. If seq(x) = seq(y) and T[x − 1] = T[y − 1], then report
(x, y) as an answer.
The trie can be created in time linear to the size of Cfk . Note that the key step
in the creation of a trie is the sorting of k-modified suffixes. To do it efficiently,
we map each k-modified suffix to the lexicographic rank of the suffix obtained
by removing all characters (from left) until its last # symbol. Using this as the
key, the k-modified suffixes can be sorted via integer sorting (refer to Sect. 2.2).
The second step of extracting answers can also be implemented using the exact
same procedure described in Sect. 4.1. However, the problem with this approach
is that, a pair (x, y) can get reported more than once, although only once per
set. In the worst case, an answer can get reported H k times. The resulting time
complexity is therefore O((N + occ)H k ).

Improving the Run Time Complexity. To achieve the claimed O(N H k +

occ) run time, we need to ensure that each output (x, y) is reported exactly
once. For this, we explore Property 2 as follows: while processing a pair of two
k-modified suffixes TΔ [x..] and TΔ [y..] under the subtree of some marked node,
report (x, y) as an answer iff
1. lcp(TΔ [x..], TΔ [y..]) = lcpk (T[x..], T[y..]).
2. seq(x) = seq(y)
3. T[x − 1] = T[y − 1]
From Property 2, for a pair (x, y), there will be only one pair of k-modified
suffixes satisfying this condition (1). The following is unique to that pair: T[x +
l − 1] = T[y + l − 1] for all l ∈ Δ. Therefore, the task can be executed efficiently
by processing the set of k-modified suffixes in the subtree of each marked node
w as follows:
1. Partition them into (at most) Σ + 1 buckets based on the previous character
as in Sect. 4.1.
2. Partition the k-modified suffixes in each bucket into (at most) |Σ|k sub-
buckets based on the sequence of k characters that were originally at the
positions in Δ. Each sub-bucket is therefore associated with a unique string
of length (1 + k): the (previous) character corresponding to the bucket in
which it belongs to, followed by the sequence of k characters at the positions
in Δ.
3. Within each sub-bucket, sort the k-modified suffixes based on the identifier
of the sequence to which it belongs.
4. Finally, for each TΔ [x..], we visit each sub-bucket and find answers of the
form (x, ·) as follows: let c0 , c1 , c2 , . . . , ck be the sequence of (1 + k) characters
corresponding to a sub-bucket. If c0 = T[x − 1] or T[x − 1] is some $i symbol
and ct = T[x + t − 1] for t = 1, 2, . . . k, then for all entries TΔ [y..] in the
sub-bucket with seq(x) = seq(y), report (x, y) as an answer. Notice that the
entries within a sub-bucket are sorted according to the sequence identifier.
Therefore, all entries with seq(x) = seq(y) comes together as a contiguous
chunk, which can be easily skipped.
12 S.V. Thankachan et al.

Analysis: The overall time for implementing the first three steps is O(N H k )
and final step takes O(N H k |Σ|k + occ) time. Therefore, total time complexity
is O(N H k + occ), assuming k and |Σ| are constants.

Theorem 2. Problem 1 can be solved in O(N ) space and O(N logk N + occ)
expected time, assuming k and |Σ| are constants.

Note: If the hamming distance between two reads is < k, then our algorithm
will not output them as an answer. To capture such answers, we shall run the
algorithm for all numbers of mismatches starting from 0 up to k. The run-time
remains the same.

5 Preliminary Experiments

We have implemented our algorithm using the C++11 standard. As noted ear-
lier, we use SA, LCP array, and RMQ data structures to simulate all the opera-
tions on the suffix trees. We construct the suffix array for the set of the sequences
D using the libdivsufsort library [12]. We build ISA corresponding to SA by a
single pass over SA. Construction of LCP array and RMQ tables are based on the
implementations in the SDSL library [5]. For the construction of the LCP and
RMQ tables, we use Kasai et al.’s algorithm [7] and Bender-Farach’s algorithm
[2], respectively. Although the SDSL library supports bit compression techniques
to reduce the size of the tables and arrays in exchange for relatively longer time
to answer queries, we do not compress these data structures. Instead, we use
32-bit integers both for indices and prefix lengths.
Note that our algorithm does not require the internal nodes to be processed
in any specific order. Therefore, we do not need to construct parent-child links
or suffix links, which are typically present in a suffix tree. We only need a list
of all the internal nodes, their string depths and the corresponding suffixes.
We represent the internal node u by the tuple (sp(u), ep(u), string-depth), where
sp(u) and ep(u) are the left and right indices in SA anchoring the range of suffixes
having path(u) as their prefix.
We conducted our preliminary experiments on a system with Intel Xeon E5-
2660 CPU having 10 cores and 64 GB RAM. We created a 2,049,118 read input
derived from the RNA-Seq dataset with accession number SRX011546, from the
NCBI SRA repository. This dataset corresponds to an Illumina sequencing of
Human CD4 T cells, with 45bp reads. Table 1 shows the run-time results for this
dataset for different numbers of mismatches allowed k. The length threshold φ
is chosen as 25.
The run-time complexity for generating all valid occurrences increases expo-
nentially with k. As shown in Table 1, this behavior is observed for smaller
values of k. However, for larger values of k, the run-time escalation is signifi-
cantly smaller than the exponential dependence predicted by the theory. This is
probably because as the matches extend towards the end of the sequences, the
number of compact tries that need to be generated is limited.
Another random document with
no related content on Scribd:
evening; I was afraid your mother was not able so you could come.”
“She is some better. She thinks and wished me to come, as she says I have
stayed at home very much of late on her account. I shall not stay very long
tonight, as I think she did not feel as well as usual.”
The young men went out on the piazza and were viewing the scenery by
moonlight, when who should they espy but Minnie and Ralph coming towards
them.
As they came up to them Minnie said, “Why brother, where is cousin?”
“She is with mother. I came to find you. Where have you been all this time?”
“You could not have looked very sharply, or you would have found me, as we
have been following you for some time and wondering where cousin was,” said
Minnie.
“Come, Mr. Burton, we will go and find her,” said Warren, leading the way
through the company to where Nettie and her aunt sat chatting pleasantly.
As they came to them Paul said, “Miss Spaulding, please favor me with your
company for a waltz?”
“Please excuse me, I never dance,” said Nettie, smiling.
“We will promenade then, if you wish. I do not care to dance either,” answered
Paul.
She took his arm, and as they walked along comments of praise were lavished
upon them, as they made a splendid-looking couple; and many of the company
saw at a glance that the young man loved to be in the company of the strange
young lady; and many a young lady there knew that he loved to dance, but
preferred the company of the lady by his side. Many envious glances were
given Nettie that evening, but she appeared not to notice them. She used them
all alike, and, when not in company with Paul, she would seek the company of
her aunt and uncle and look on and see the others enjoy themselves. “I do not
care to dance,” she would tell all who asked for her company. “I cannot enjoy
dancing,” she would say to her aunt, when she urged her to dance.
“No, no, I can not,” she would say.
It was not because she could not dance. It was because of a request of her
father, who was lying in his grave, and of her mother who was far away at
home.
How many young people of today scarcely wait till the green sod grows over
the grave of some beloved form, before they are away to some ball or place of
amusement? Such is progression.
It was getting quite late and Paul came to Nettie and said, “Miss Spaulding,
accept my company, please, for a promenade on the piazza. The moon is
spreading its rays beautifully and the evening is delightful.”
She took his arm and they walked quietly out under the trailing vines of
myrtle, which were trained to droop from the eaves of the old farm-house.
They came to an old-fashioned settee that was enfolded in the drooping vines
and formed an arbor. Here they sat down. Soon Paul said, “Miss Spaulding,
have you been down to the lake since Monday?”
His companion blushed deeply as she answered. “I have not, sir; you must
have heard all I said, did you not? I was very lonely that day—my poor mother
far away and I alone here. My cousins are very kind to me, very kind indeed, or
I do not know what I should do.”
“Will you accept the friendship of a stranger? As you know but little about me
that is all I will ask now. I never saw a lady in all my wanderings who ever drew
such words of acknowledgment from me before. All I ask is friendship, and
when you know me better perhaps I shall ask you for this little hand.”
He gently raised her hand to his lips as he was speaking.
She drew it quickly from him saying: “Sir, please pardon me if I have given you
occasion to make the declaration. The truth we should tell at all times; perhaps
you think me rich; if so, you are mistaken. I am very poor. Such as you needs
not the friendship of one beneath him.”
Truth and honesty shone in her dark, brown eyes as she turned her head away
to hide the gathering tears. It pained her very much to tell him whom she
loved. She had been taught to shun deceitfulness, and she thought it decisively
her duty to tell him she was poor, no matter how it pained her to do so. She
spoke deliberately, but in a dejected manner. She was pale, with a faint flush on
her cheeks that was drawn there by the enthusiasm she was forced to exercise.
“Nettie, darling, you do not know me. It is not wealth I wish. It is this little
being by my side. She is rich in voice, rich in beauty, and richer still in mind.
Do not say wealth to me again—it hurts my feelings.”
As he spoke he gently drew the little form nearer to him and rested her head
on his great, manly breast.
“Only four days have I known you, yet it seems to me a life time.”
Nettie quickly arose saying: “Please, sir, say no more to me; always remember
me as your true friend, one who will not do you an unkindness. Never say
aught of this meeting to anyone for my sake and for yours, and in the future if
you prove faithful to me I am yours.”
She turned and fled away, leaving him sitting in the twilight deeply touched.
How long he had been there he knew not. Warren Hilton’s voice brought him
to his senses as he said, “Paul, where is my cousin? I have not seen her since
you came out together.”
“Oh, Warren! I do not know; she abruptly left me here, and how long I have
been here I know not. Oh, I have stayed too long. I must surely go home.”
He quickly arose, and he looked so sad Warren really pitied him as he said,
“Why, Paul, are you sick?”
“Oh, no,” answered Paul; “only sorry I have stayed from home so long.”
“I hope you have not been unkind to my little cousin,” said Warren changing
the subject, as he thought Paul was really thinking of his mother.
“Been unkind to her? been unkind to your cousin?” said Paul, looking Warren
squarely in the face; “I would sooner cut my right hand off than say one word
to offend that lovely little girl.”
Warren saw he was deeply troubled as he answered, “Paul, what then is the
matter?”
“I cannot tell you; go find your cousin. Perhaps she will tell you.”
Paul’s voice trembled, and Warren readily guessed the cause, as he thought
Paul had sued for the hand of his cousin and had been refused. He went to
find Nettie and he thought she would readily tell him all he wished to know.
He looked, but could not find her anywhere among the company. At last he
found Minnie and asked her where Nettie was.
“I have not seen her for a long time; I saw her last with Mr. Burton.”
“She is not with him now and has not been for some time. I wish you would
go up to her room and see if she is there,” said Warren, “I fear something is
the matter with her.”
Minnie ran softly upstairs to her room. She heard someone walking to and fro
as if in a hurry. She gently rapped at the door and a trembling voice bid her
come in.
“You know you are always welcome, Minnie.”
Not heeding Nettie’s words Minnie said, “why did you come up here? Warren
missed you and sent me to find you. Why, Nettie, where are you going? I see
you have been packing your trunk.”
 CHAPTER IV.

“I hope you are not going home?”

“Yes, cousin, I shall go tomorrow. I wish you would bear the intelligence to
your father and mother and entreat them to let Cousin Warren take me to the
village in time to take the coach for home tomorrow. I do not care to ask him,
as he will question me. Mother will be looking for me in a day or so and I
concluded to go tomorrow.”
She gave her hand to her cousin, saying: “We will go down to the hall now or
the company may think it strange we are both gone.”
They went along the corridor as placidly as if nothing had transpired to mar
the pleasure of the evening.
They came to where Warren stood toying with the tassel of the window
curtain and looking out into the moonlight with deeply-troubled thoughts.
“Where have you been, you little rogue? we have been looking for you for
some time.”
“She has been packing her trunk to go home tomorrow morning: I cannot get
her to stay any longer,” said Minnie.
“What has caused you to make such a quick decision? I supposed you were
going to stay with us two or three weeks. Something has offended you I fear,
or you would not decide so quickly to go home.”
“No, Warren, nothing has happened of any account: please don’t scold me,”
said Nettie sorrowfully, as she was nearly overcome with the burden on her
mind.
She turned her head away to hide the tears from prying eyes. She turned to go
when Warren said, “Please excuse my last words, cousin, I did not wish to
scold you. See! yonder comes Mr. Burton. He is coming this way.”
He was calm but pale. As he drew near to them he said, “Mr. Hilton, get my
hat please; it is time for me to be going.”
As Warren left to do his bidding Paul said, “Miss Spaulding, if I have said
aught to offend you, pardon me. As God is my witness, what I have told you is
the truth. I will do as you have bidden me to do, and I ask in return to
remember me some times when alone.”
He gently pressed the little hand he was holding.
“Goodbye, and may God bless you forever,” said Paul solemnly.
Soon Warren came with the hat, and Paul taking it bid them all good night and
went homeward in a sad frame of mind.
One hope she had given him, viz: “If you prove faithful to me in the future I
am yours.”
These words cheered him, and he fully resolved to be true to her until death.
“What can be her object. Can it be she thinks she is not good enough for me
financially?” thus murmured Paul until he reached home.
He found his mother sitting up. She had been having a serious spell of heart
disease and dared not lie down. As he entered the room she was sitting in she
said, “My son, why did you come so soon? I did not expect you for some time
yet.”
“It is nearly twelve, mother, and I am sorry I stayed so long. You have been
sick, and are now, only wishing to keep me in ignorance of how bad you really
are. You look very ill mother. Why do you sit up so long?” asked her son,
bending over her and pressing a kiss on her fair brow.
“My son,” answered his mother, “I have the heart disease, and I fear you will
soon have no mother. I see it is growing worse with me with every attack, as I
cannot lie down after one now.”
“Oh! mother, do not speak so sadly. Shall I go for a doctor tonight?”
“It would do me no good. I have tried the best-skilled physicians there are on
the continent and they unite in saying I must be kept quiet or I will some day
be no more. I have prayed that I might live to see you grown to manhood, and
that prayer is answered and now I am willing to go when God sees fit to call
me.”
His mother was speaking in a sorrowful tone. Paul sat like a statue, pale as
death.
“Oh! mother, it cannot be,” he spoke at last. “I can not part with you; you who
are all the companion I have on earth,” answered Paul in frightened tones.
“My son you will not miss me much when you catch that ‘little human fish’
you spoke of the other day. Oh! if you should marry her I pray she may prove
a true, honest wife to you. Then you will lead a happy life.”
“Oh, mother, may your last few words be true! Time works wonders in this
world sometimes. I hope you may live long with me, then you will see what a
dutiful son you have,” answered Paul, the tears falling thick and fast.
The nurse came in with a cup of strong tea for his mother, and Paul arose as
he said, “Take good care of my mother and I will repay you well.”
He kissed his mother again and went off to bed but not to sleep. Try as he
might no sleep came to his eyes. Early the next morning he arose, took his
shot gun and went out to see if he could kill a pheasant, to make some broth
for his mother.
The next morning after the party Mr. Hilton said to his niece as she came
down to breakfast: “You did not receive much pleasure by the party I fear,
Nettie. Warren has been telling me you wish to start for home this morning. I
hope my children have not done anything to mar your pleasure here.”
“Uncle, they have done nothing to mar my happiness,” answered Nettie, with a
dreary laugh that touched her uncle’s heart.
“Will you promise me if you and your mother ever come to want that you will
come and live with us? Our house is large and you are both welcome to its
shelter.”
Nettie went gently to him, planted a kiss on his fair, honest forehead and said,
“I promise. Never can I forget the kindness I have received at this new home,
or forget the inmates that dwell here.”
The eyes of all of her friends were filled with tears to see the sweet young girl,
who, standing smoothing her uncle’s silvery hair, was outwardly calm, but a
deep trouble was raging in her breast, as she wished to stay but could not and
did not wish to let any of her friends know the real cause.
Her aunt said, “My dear, something has transpired to make you decide so
quickly to go home. We expected you were going to remain two or three weeks
with us.”
“Nothing, auntie, only my conception to go home and surprise mamma. She
will be very delighted to hear from you all. Of course I shall tell her what a
pleasant time I have had with my cousins. It is getting late and we should be
going soon or we will not be in time to take the stage, as it leaves at ten.” So
saying she began putting on her mantle and cap. As her cousin Warren drove
up to the door with a splendid span of iron grays, he called out lightly, “All
aboard for town.”
“Auntie, are you and cousin Minnie not going to see me off for home?”
“No, my dear, we cannot go with you, as we have these rooms to make tidy.
Warren will see you safe there and in the coach, too.”
Nettie bade her friends good-bye and was still lingering at the door, as she was
loath to leave her new found home.
“I will come back here some day perhaps, and then I will stay longer; or long
enough to make you wish there never more could come a Nettie Spaulding to
trouble you,” said Nettie, feigning a laugh.
“Never you need be afraid of that,” answered her uncle, “come and see us and
stay as long as ever you can. We will be most happy to see you.”
“Thank you, uncle, I will return soon no doubt.”
She tripped lightly out, and Warren handed her into the buggy, and soon the
two cousins went from that farmhouse in a very sad mood, as Nettie was
leaving her new found friends to go back to the great busy city to live within
herself, as her old associates avoided her, or she avoided them, as she could
not meet them as of old.
Warren was sad, as he did not wish his little cousin to leave them. She was like
a sunbeam in the dear old home, and he had taken great pleasure in getting the
two young people together who he thought were best suited to each other.
Now his pleasures were ended, as his cousin was going home.
“If she was not my cousin,” he would say to himself, “I would try to win her
affections, but that word cousin casts all into oblivion as far as I am
concerned.”
As they were driving over the rough country roads, Warren said, “Little cousin,
there is something wrong or you would not be leaving us so soon; is there
not?”
“Warren,” she said, bursting into tears, “God alone knows the misery I have
endured since last evening. You say you are my friend; I believe you, as you
seem to take great interest in my welfare. I am going home to live like a
hermit, in a great city. As such always think of me. I would like to stay, but it
can not be,” she exclaimed passionately.
“He is rich and I am poor. I can not stay and be a temptation to one who is
dearer to me than life. If he proves true then all is well, if not, then God pity
me.”
Warren was listening to her passionate words, while tears stood in his honest
blue eyes as he said, “Paul Burton is a man of honor. If he told my little cousin
he loved her it is the truth, as I have known him for many years, or ever since I
can remember a playmate, and I never have caught him in a lie.”
Nettie was weeping violently as she said, “Please write to me often, and write
all the news about him, but do not tell him one word about me. If he really
loves me he will find me, if not, it is better as it is.”
She spoke sadly.
“I will do as you have bidden me,” said Warren, “and prove to you that I am a
true friend.”
Suddenly the crack of a gun was heard. The horses sprang forward and nearly
threw the young couple out of the buggy.
“I wonder who is out sporting so early this morning,” said Nettie.
“It sounds like Paul’s gun,” said Warren, as he gently drew up the reins of the
horses and brought them to a walk.
“I wonder how Paul’s mother is this morning. He said she was not very well
last night. Perhaps he is out to kill something for her.”
“Has his mother been sick very long?” asked Nettie.
“She is a tall, frail woman, and she has very bad spells. Some people say she
has heart disease,” said Warren.
“I am very sorry indeed. It would be very sad for him to have his mother
taken from him. I really hope to hear when you write that she is better.”
They went slowly up to the little village hotel. The stage was about to start.
As Warren handed her down he said, “Do not forget to write me all your
troubles, cousin, and I will write you the news. I will give the same injunction
to come and live with us as father did.”
“Thank you for your kindness. I shall never forget you or the dear ones I left
in my new home by the lake,” answered Nettie.
“Have you no word for Paul?”
“Yes, cousin, tell him good bye, to be upright and honest in all his endeavors,
and God will deal justly by him. Good bye, cousin,” said Nettie.
As she took a seat in the stage she peeped out of the window and said, “Write
me often, and please send me the village newspaper if you do not think I am
asking too much. I will send the change when I arrive home.”
“I will go and order it sent you so you will get it next week,” said Warren.
The stage started on its long journey to the city, bearing one sad little being on
her way for home.
How happy it makes one feel to unburden a troubled mind to a true friend,
and it seems to make the heart lighter to have words of consolation given in
the hour of trial from a true, loving friend. Many a young person and many
aged ones can bring back to memory the same solemn fact.
Thus it was with Nettie as she went homeward. Warren’s kind words ever rang
in her ears: “He is a man of honor; if he told my little cousin he loved her it is
the truth.”
How many times in the future did she think of them and draw consolation
from them.
Warren watched the stage that bore his cousin homeward until it was out of
sight, then started homeward at a brisk pace.
He had not gone but a few miles when he overtook Paul returning from
hunting. On his shoulder hung several pheasants.
Warren brought his horses to a halt as he said, “Take a seat by my side, Paul, it
is better to ride than to walk. Are you not tired? You must have gone out early
this morning, as I heard the report of your gun when I went to town.”
“Yes,” answered his companion getting into the buggy, “I came out very early,
as mother is not as well as usual and I thought some wild food would be good
for her. I fear my mother is not long for this world, as she is failing every day. I
sent Pompey for the doctor this morning, but mother says it will do no good,
as she is past cure. Oh, Warren, I do not know what to do or where to turn,
for I am in deep trouble. Why don’t you come over oftener and stay some
night with us?”
“Would your mother be willing? She is so delicate about company,” answered
Warren.
“She would be very happy indeed to see me have company,” said Paul.
“I will come over in a day or so,” said Warren.
“Please do, Warren, in an hour of need, as I am very lonely—mother sick, and
she is my only companion except the servants.”
They came to a cross road that was nearer for Paul to reach home and he
sprang lightly out and ran swiftly home with his game.
Nettie’s homeward journey came to an end in due season, nothing happening
of any account worth mentioning. As she came sooner than her mother
expected her she was surprised to see her child back again.
In less than a week after greetings were exchanged and many questions asked
about distant friends the mother said, “Why, Nettie my child, why did you not
stay longer? I did not expect you for two weeks at least.”
 CHAPTER V.

“Oh, mother, I could not stay away any longer from you. It seemed a long
time to me.”
“Why, my child, in your letter you said you was happy and would stay two
weeks, as your uncle and cousins would not take ‘no’ for an answer and wished
we should come and live with them; and I was nearly making up my mind to
go up there for a while and see the country. Perhaps it would be agreeable to
my health.”
The mother was viewing her child critically while speaking. Noticing Nettie’s
face changing from a bright crimson hue to a pale color, and not answering
her, she said, “Has my little girl quarreled with anyone out there and come
home angry?”
“No, no mother,” answered Nettie.
“It is worse than that, mother; I will tell you all, as no true mother would
advise her child to do wrong. I will tell you all, but do not think for one
moment I was telling you an untruth when I told you I could not stay longer. I
could not under the circumstances, and it seemed a long time to leave you here
alone. Well, mama, now for my story: My cousins made a party for me last
Friday evening, as you know Friday was my birthday, and invited all the young
people in that vicinity, and among them was a rich young man, highly
esteemed for his true manliness. Cousin Warren says he has known him ever
since he can remember, as they have grown up and trudged to school together,
and says he never caught him in a lie. That is saying a good deal about him.
Well, the Monday before the party cousin went to town and I went down to a
beautiful lake on uncle’s farm to gather flowers. I sat down on the brink of the
lake and some of the flowers fell into the water. I was wondering how to get
them when the same young man spoken of came and fished them out. He
gave me his card, and the night of the party he told me the same old, old
story.”
“What did you tell him Nettie?”
“Oh, nothing in particular. The most I told him at last was if he proved true to
me in the future I was his.”
“Nettie, dear, do you really love him?”
“Yes mother, with my whole heart. But I have run away and if he really loves
me he will hunt me up,” said Nettie, her face beaming with smiles.
“Nettie,” said her mother, “pray what is the name of your admirer?”
“Paul Burton, of Pine Island. The name was given to the farm many years ago.
It is a beautiful farm enclosed by the forest, and there is a little lake on it; and
in the center pine trees are growing. I was out with cousin and he took me by
there.”
Nettie was speaking with enthusiasm and hearing her mother repeating the
name she turned and noticing her mother’s pale face said, “Mother, what ails
you? Are you sick?”
“No, Nettie, the name sounds familiar. What kind of a looking man is this Paul
Burton, and what is his age?” asked the mother.
“He is tall—about six feet—well proportioned, his eyes are dark blue, and he
has auburn hair, and is a picture to behold,” answered Nettie.
“Blue eyes and auburn hair; did his hair curl?” asked her mother.
“Yes,” answered Nettie, “and he is about twenty-two or three. He lives alone
with his mother, who is a frail, sickly woman.”
“Did you ever see her?”
“No, mother, but cousin says she is tall and dark complexioned, with black
eyes, and her given name is Margaret or Margretia, I do not know which.”
“It is the same woman and must be their son. Oh, my God! why have I come
to this?” exclaimed Mrs. Spaulding.
“Why, mother, what is the matter, and who are you referring to?” asked Nettie,
noticing her mother’s pale face.
“My child, one you never saw—and I hope you may never meet him or any of
his descendants.”
“Why, mother. His descendants should not be cruelly judged by his conduct.
You speak as though he had been guilty of some great criminal act. I do not
see what he has to do with Paul Burton, the young man I was speaking of,”
said Nettie, turning and looking out of the window.
“If I had known it would have troubled you, mother, I would not have told
you anything about him. You seemed so anxious to know why I returned so
soon I thought it proper to tell you all. The young man was supposed rich and
I was a poor girl with only my good name to sustain. I deemed it best to try
his love. If he loves me sincerely he will find me; if he does not, it is better I
should be far away. Do you not think my act justifiable, mother?”
“Yes, my child, you did what is right and proper, and I am glad you came
home, and I hope my conjecture is not true,” answered the mother
sorrowfully.
Nettie went to her mother and pressed a kiss on her pure fair brow. She had
passed through many severe trials, yet she remained beautiful—only a trifle
pale. Time made little impression upon the fair form of the once beautiful
Minnie Hilton, one of old England’s fair daughters.
“Nettie, I have a long story to tell you. It might prove a good lesson to you in
the future, as you are young and inexperienced in this world of sunshine and
shadow, and you may draw conclusion from the story.
“My child. I hope you will not have to endure the troubles and sorrows like
the lady of whom I am going to speak.”
“Oh! mama, do tell me now, as I am anxious to hear it. I am sure it is a
warning to me,” said Nettie, tapping the velvety cheek of her mother.
“Well, Nettie, many years ago in England there once stood a neat cottage
surrounded by a group of beautiful trees, and just within hearing of the big
bell in London. What a happy little home it was before the revolution broke
out in this country. England was all confusion, especially among the second
and lower classes of people. The inmates of that little cottage numbered four
—father, mother, a lovely girl of eighteen summers, and a lad of sixteen, as
honest a boy as ever lived and a kinder heart never beat today. Well, the father
had to help to fill the ranks of England’s army and came over here to fight for
King George. How noble and manly he looked in his red coat as he mounted
his coal black steed. He made a fearless and brave soldier, as many of his
comrades testified on their return home. But he who kissed his wife and
children an affectionate farewell never returned to receive their welcome
embrace of joy as did many of his fellow soldiers. As the news spread quickly
over the old domain that the battle of Bunker Hill had been fought many tears
fell for the fallen soldiers who fell in that sad fray.
“Sad was the news indeed, to hear that the father of this happy household was
no more, during the intervening term of his going away and time of his death.
The daughter of this family was the fairest in England at that time. Her fame
for beauty rang far and near. One day in summer when the commons were
robed in green, besprinkled with buttercups and daisies, this young lady for a
little pleasure rambled over the green, picking the flowers and thinking of her
father who then was far away in the battle fighting for his king, when close to
her she espied a large stray sheep of the masculine gender. He had probably
broken out from its owner’s enclosure and was wandering over the commons.
As soon as he espied the lady he came toward her with his head bent to the
ground, and the lady gave a scream and was running toward the hedge fence
of thorns; and just as the sheep was about to strike her a young man rode
rapidly between them, striking the sheep with a heavy loaded whip, which
felled him to the ground as though dead. The gentleman sprang lightly from
his horse and picked up the inanimate form of the lady, as if she was an infant,
and bore her to a cottage near by, and by the aid of spirits she soon returned
to consciousness. She had swooned with fright and had fallen, hitting her head
lightly on a rock, cutting a cruel wound which bled profusely. The young man
saw the blood and he only had thought for the fair young form as he quickly
bore it to a friendly shelter, letting his horse roam at will. The lady was too
weak to walk, so the gentleman went home and took his father’s carriage and
took her home; as he called every day for several days to see how his patient
was getting along he grew deeply in favor with the little family of the cottage.
The young lady looked for his coming and was deeply grieved when she
learned who he was, for he was the son of a baronet, a gentleman of note
among the upper classes of people. He was a lovely young man, and one
beloved by old and young throughout the community. He called often at the
cottage. None of the inmates could tell him to come no more, as he was both
manly and honest, and with each day he grew more enamored with the little
lady he had saved from a cruel death. How time flew away! Soon his father,
who was not noted for kindness, began to notice his son’s movements, and it
soon became known to him where he wandered. As his son was of age he had
him sent off to the war, as he would then get over his love passion for the little
cottager, as he called the little lady his son admired. Sad was the last meeting of
the young couple, as he came to bid her farewell. Many were the promises
given each to prove faithful until death. Then another blow was given that
household, as he in his red coat rode away leaving his promised bride to
mourn the loss of one she deeply loved. Soon after came the news of the
battle of Bunker Hill, and the father of this little cottage was no more. Deeply
mourned the inmates for the friend and father, and also for the absence of one
who seemed a true friend to all; but he was the King’s subject and had to go
and leave a lover behind him to mourn his absence, as many over our land
today have done, and how sorrowful the earth seems to the ones left behind.
“The young man went away in hopes of a speedy return, but what a sad
delusion! One year passed, then a second, and a letter came to the loved one
far over the deep that her lover was slain. How deeply that little girl mourned
for her supposed dead lover no human tongue can tell, and as time flew away
many were the changes with the inmates of that cottage. Finally in time the
mother concluded to remove her family over here to America. She wished to
view the resting place of her beloved companion, and when the shadows of
death came to her weary soul her form might lie in the same soil, beside her
husband. Cold and stormy was the day when the noble ship set sail that bore
on its bosom the widow and her children. It was the following autumn when
they landed in New York, six weeks after setting sail. The widow rented a little
cottage and made the place her future home. Her children both had grown to
manhood and womanhood, and having a good education managed to maintain
themselves respectably.”
“Oh, mama, you did not tell me whether the young lady ever heard definitely
about her lover’s death,” said Nettie, breaking in on her mother’s narrative.
“My child,” answered her mother, “she left word with some of her friends that
if any news came to them concerning him they should write to her
immediately. She received only one letter bearing news of him. It said he had
returned before the widow had left England. It was reported by the young
man’s proud family that he was dead, for they knew their son’s disposition
would be to fulfill his promise to the ‘little cottager,’ as they called his
promised bride. They were bound it never should be. At the last meeting
before he went to war he frankly told his father he should marry the girl on his
return home; he might disinherit him if he chose—he would have a few
shillings of his own and he would take his wife over to the new land he was
going to fight on. This exasperated the father to such an extent that he
brought his fist down on the table and swore an oath it never should be. The
son did seem not to heed his father’s words, as he was sure the lady would
prove faithful and he would soon return and claim her, in spite of all earthly
beings. The lady watched and waited until nearly autumn for a letter from her
loved one which never came. During that time she received news of her lover’s
death, then broken hearted, she urged her mother to leave the place which,
with each day, brought memories of the two loved ones who never would
tread o’er its well-remembered threshold. The young lady lived single eight
years. She employed the time in teaching primary school in the suburbs of the
city of New York. Time, the great healer, brought consolation to the wounded
heart. At last she accepted the hand of a young merchant. The old love was
buried beneath the new, but never forgotten. The young lady’s brother learned
the trade of a mechanic but did not like it very well, and like Washington
turned to agriculture as soon as he earned money enough to buy a farm. The
widow often went to view the grave of her beloved companion, and when her
life on earth was ended her children laid her silent form beside him she loved.”
“Why, mamma!” exclaimed Nettie, the tears trinkling down her fair cheeks, “it
was my own grandmamma, as I can just remember when papa and Uncle John
took her deceased form away and did not get back for a long time,” said
Nettie, speaking slowly.
“Why, mamma, can the young lady you have been speaking about be
yourself ?”
“Yes, my daughter, it is the same.”
Nettie was standing by her mother’s chair, stroking the fair brow of her only
parent, deeply thinking over all she had been told.
Soon she said, “Mamma, it was very sad indeed, but I do not see why it should
be a warning to me.”
 CHAPTER VI.

“My child,” answered the mother sorrowfully, “you will be surprised when I
state that the young man did go home and did not try to find where we
removed to, but soon became acquainted with a lady of high standing and
married her. They came over here, bought a large tract of land somewhere in
the New England states. Your father and I met him once on board a vessel
lying in harbor. Your papa, dear soul, knew of my first love and also the name
of his predecessor, and getting an introduction to him made himself very
inquisitive, as he found he was the same person and had been married some
twelve years. You see how deceitful some people are in this world. It is a good
saying and a true one, too, that ‘sometimes you think a friend you’ve got until
trial proves you have him not.’ Thus it proved out to be to me.”
“Mamma,” answered Nettie, “I am a trusting spirit. It might not have been all
his fault. He might have been deceived, as the story circulated by the family
deceived you. Perhaps he did try to find you but could not get any clue to your
whereabouts, as money sometimes will do a great deal in the way of bribing
people.”
“Well. Nettie, time will prove all things. As it is said, ‘Right conquers might.’”
“Mamma, what is the name of the once young man you have been speaking
of ?”
The mother looked sadly up in her daughter’s fair face as she answered, “Paul
Burton; with manly form, blue eyes, and hair the color of your admirer’s.
Nettie, I am so glad you had sense enough to come home, as it is my
conjecture that this young man is the son of him I have told you of. God grant
it may not be!” said the mother fervently. “I do not wish my child to be
deceived as I have been.”
“Your father and I lived very happy through the years of our married life. No
shadows came to mar the horizon of our union until he became bankrupt
through a person we supposed our friend. Poor soul, like many others before
him he could not stand the crash in his financial affairs, and soon after died. It
was a sad blow to me as I loved him fully as well as I ever did Paul Burton, the
baronet’s son.”
“Poor mamma!” exclaimed Nettie passionately. “God must have willed it to be
so. I love this strange young man and it was very hard for me to come home
and leave him whom I loved so fondly, but my English pride bade me come
home.”
“And I hope God will deal justly by us all. We must trust to Providence and
wait and see what time will bring in the future.”
“Oh mamma dear, I cannot believe Paul is false. Oh, no, no, it cannot be!”
exclaimed Nettie passionately.
“May God be merciful to me.”
“My child. God doeth all things. We can trust to Providence and all yet may be
well. This is a world of trouble and sorrow to us poor mortals and what falls
to our lot we must endure patiently, for what is to be will be, in spite of all
human aid. I sincerely hope for the happiness of my only child,” said the
mother, pressing Nettie fondly to her breast.
Here I will leave them bemoaning their fate, and return to Paul, who, on
returning home, found his mother very ill. She gradually grew worse day by
day. All medical skill was of no avail and they could not restore her to ordinary
health. Time passed drearily at that once pleasant home. Paul, sad hearted,
went about the house as one in a dream, never speaking to any one except to
give orders to servants and inquire about his mother, whom he loved more
fondly than ever. He knew she would soon leave him, and it grieved him very
much to see her sad, pale face as she would look fondly at him and say, “I will
soon be at rest—free from all earthly trouble.”
She lingered through the fall and long dreary winter months, and as the buds
came on the trees in the following spring she breathed her last, while lying in
the arms of her affectionate son. Sad was the scene, to see the young man
fondly clasp his mother to his breast while tears fell like rain, on the sad, silent
face.
A few moments before she died she called for her son, and when he did not
come immediately she said, “Must I die and not tell him? I ought to have done
it before now. Oh, where is my boy? Will he come soon?” she asked faintly,
turning her face to the wall.
Pompey, hearing her call, went into the room in time to hear her last words.
He went to her bedside as he said, “Paul will come soon. He went down the
lane to see if the doctor is coming. I’s sent for him and he will come very fast
when he hears de news. Missus, I’s been berry kind and obedient to you, ain’t
I? I’s lived with you ever since Paul was a little chick. Anything you want me to
tell him, Missus?”
“Yes, my faithful man, you know the whole history of my life, and when I am
gone tell him not to censure his father as it was all my fault—his leaving home;
but make my sin as light as you can. There is a little tin box in the garret that
will tell him all he wishes to know.”
She nearly held out her hand to the faithful man, saying, “It’s very hard
bidding you goodbye, Pompey, and may God bless you forever!”
She whispered the last words, and as her son came into the room her eyes
brightened for a moment and she tried to speak to him but could not. Her
breath grew shorter and shorter with each moment, and soon she was no
more.
They laid her beneath the weeping willow tree and at her grave the son placed
a neat monument in memory of her who reared him to manhood. Sad and
dreary was that house to Paul. The sunshine had fled and only shadows
remained. No mother now to kiss him good night; no father to bear with him
this sorrow, and the only being he loved beside his mother was gone, he knew
not where. The only friends that deeply sympathized with him except the
servants were John Hilton’s family, especially Warren, who was there night and
day and kept Paul company through this sad affliction. When this kind
companion went home Paul could not reconcile himself to stay in the house
where once was life and joy for him.
“I cannot stay here; I must go somewhere; there is no comfort for me on this
earth. Oh, why did I live to see this trouble!”
Thus he would talk to Warren Hilton when they were alone.
“Why do you not go away from here for a while? The servants can look after
the farm, and I will run over now and then to see how they get along. You can
write me and you can hear all about them. You can go down to the city of
New York, or anywhere else you choose. Something may change for the best. I