skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Method and apparatus for biological sequence comparison

Abstract

A method and apparatus are disclosed for comparing biological sequences from a known source of sequences, with a subject (query) sequence. The apparatus takes as input a set of target similarity levels (such as evolutionary distances in units of PAM), and finds all fragments of known sequences that are similar to the subject sequence at each target similarity level, and are long enough to be statistically significant. The invention device filters out fragments from the known sequences that are too short, or have a lower average similarity to the subject sequence than is required by each target similarity level. The subject sequence is then compared only to the remaining known sequences to find the best matches. The filtering member divides the subject sequence into overlapping blocks, each block being sufficiently large to contain a minimum-length alignment from a known sequence. For each block, the filter member compares the block with every possible short fragment in the known sequences and determines a best match for each comparison. The determined set of short fragment best matches for the block provide an upper threshold on alignment values. Regions of a certain length from the known sequences that have a mean alignment value uppermore » threshold greater than a target unit score are concatenated to form a union. The current block is compared to the union and provides an indication of best local alignment with the subject sequence. 5 figs.« less

Inventors:
;
Publication Date:
Research Org.:
Cold Spring Harbor Lab
Sponsoring Org.:
USDOE, Washington, DC (United States); National Insts. of Health, Bethesda, MD (United States)
OSTI Identifier:
563676
Patent Number(s):
US 5,701,256/A/
Application Number:
PAN: 8-455,654; CNN: Grant 1R01 HG0020301A1
Assignee:
Cold Spring Harbor Lab., NY (United States) PTO; SCA: 550200; PA: EDB-98:016256; SN: 98001896303
DOE Contract Number:  
FG02-91ER61190
Resource Type:
Patent
Resource Relation:
Other Information: PBD: 23 Dec 1997
Country of Publication:
United States
Language:
English
Subject:
55 BIOLOGY AND MEDICINE, BASIC STUDIES; DNA SEQUENCING; DATA ANALYSIS; MOLECULAR BIOLOGY; MOLECULES; MOLECULAR WEIGHT; DNA

Citation Formats

Marr, T.G., and Chang, W.I. Method and apparatus for biological sequence comparison. United States: N. p., 1997. Web.
Marr, T.G., & Chang, W.I. Method and apparatus for biological sequence comparison. United States.
Marr, T.G., and Chang, W.I. Tue . "Method and apparatus for biological sequence comparison". United States.
@article{osti_563676,
title = {Method and apparatus for biological sequence comparison},
author = {Marr, T.G. and Chang, W.I.},
abstractNote = {A method and apparatus are disclosed for comparing biological sequences from a known source of sequences, with a subject (query) sequence. The apparatus takes as input a set of target similarity levels (such as evolutionary distances in units of PAM), and finds all fragments of known sequences that are similar to the subject sequence at each target similarity level, and are long enough to be statistically significant. The invention device filters out fragments from the known sequences that are too short, or have a lower average similarity to the subject sequence than is required by each target similarity level. The subject sequence is then compared only to the remaining known sequences to find the best matches. The filtering member divides the subject sequence into overlapping blocks, each block being sufficiently large to contain a minimum-length alignment from a known sequence. For each block, the filter member compares the block with every possible short fragment in the known sequences and determines a best match for each comparison. The determined set of short fragment best matches for the block provide an upper threshold on alignment values. Regions of a certain length from the known sequences that have a mean alignment value upper threshold greater than a target unit score are concatenated to form a union. The current block is compared to the union and provides an indication of best local alignment with the subject sequence. 5 figs.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {1997},
month = {12}
}