TY - CHAP

T1 - Sequencing by Hybridization in Few Rounds

AU - Tsur, Dekel

PY - 2003/1/1

Y1 - 2003/1/1

N2 - Sequencing by Hybridization (SBH) is a method for reconstructing an unknown DNA string based on substring queries: Using hybridization experiments, one can determine for each string in a given set of strings, whether the string appears in the target string, and use this information to reconstruct the target string. We study the problem when the queries are performed in rounds, where the queries in each round depend on the answers to the queries in the previous rounds. We give an algorithm that can reconstruct almost all strings of length n using 2 rounds with O(n logα n/logα log α n) queries per round, and an algorithm that uses log α* n - Ω(1) rounds with O(n) queries per round, where a is the size of the alphabet. We also consider a variant of the problem in which for each substring query, the answer is whether the string appears once in the target, appears at least twice in the target, or does not appear in the target. For this problem, we give an algorithm that uses 3 rounds of O(n) queries. In all our algorithms, the lengths of the query strings are Θ(logα n). Our results improve the previous results of Margaritis and Skiena [17] and Frieze and Halldórsson [10].

AB - Sequencing by Hybridization (SBH) is a method for reconstructing an unknown DNA string based on substring queries: Using hybridization experiments, one can determine for each string in a given set of strings, whether the string appears in the target string, and use this information to reconstruct the target string. We study the problem when the queries are performed in rounds, where the queries in each round depend on the answers to the queries in the previous rounds. We give an algorithm that can reconstruct almost all strings of length n using 2 rounds with O(n logα n/logα log α n) queries per round, and an algorithm that uses log α* n - Ω(1) rounds with O(n) queries per round, where a is the size of the alphabet. We also consider a variant of the problem in which for each substring query, the answer is whether the string appears once in the target, appears at least twice in the target, or does not appear in the target. For this problem, we give an algorithm that uses 3 rounds of O(n) queries. In all our algorithms, the lengths of the query strings are Θ(logα n). Our results improve the previous results of Margaritis and Skiena [17] and Frieze and Halldórsson [10].

UR - http://www.scopus.com/inward/record.url?scp=0142152722&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-39658-1_46

DO - 10.1007/978-3-540-39658-1_46

M3 - Chapter

AN - SCOPUS:0142152722

SN - 3540200649

SN - 9783540200642

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 506

EP - 516

BT - Algorithms - ESA 2003

A2 - di Battista, Giuseppe

A2 - Zwick, Uri

PB - Springer Verlag

T2 - 11th Annual European Symposium on Algorithms

Y2 - 16 September 2003 through 19 September 2003

ER -