Dpuiu alignment: Difference between revisions
Jump to navigation
Jump to search
| Line 35: | Line 35: | ||
== Abstract== | == Abstract== | ||
* We present a software tool for | * We present a software tool for sequence alignment processing. | ||
* The tool filters and clusters local alignments | |||
* The tool was initially developed for finding syntenic regions in large eukariotic genomes but could also be used for analyzing rearrangements in microbial genomes. | * The tool was initially developed for finding syntenic regions in large eukariotic genomes but could also be used for analyzing rearrangements in microbial genomes. | ||
* The | * The software takes as input alignment data which is generally very fragmented (low identity) and noisy (repeats, rearrangements) and tries to filter and cluster it in | ||
* A moving window approach is used for comparing each alignment to its neighbors and decide if it agrees or not with its neighbors. The alignments which clearly disagree get discarded. | * A moving window approach is used for comparing each alignment to its neighbors and decide if it agrees or not with its neighbors. The alignments which clearly disagree get discarded. | ||
| Line 45: | Line 46: | ||
== Motivation == | == Motivation == | ||
Sequence alignment is one of the | Sequence alignment is one of the most common bioinformatics problems. | ||
Due to its complexity there is no one general solution and multiple algorithms have been implemented for aligning different data set types. There is always a tradeoff between speed and accuracy. | Due to its complexity there is no one general solution and multiple algorithms have been implemented for aligning different data set types. There is always a tradeoff between speed and accuracy. | ||
Latest revision as of 18:57, 5 October 2009
Filtering
Programs:
- filter-anc.pl -max -percent -w -W
max: max distance between the ref_intercept (default 1,000,000) percent: min percent agreement for keeping an alignment (default 25) w: min window size (default 4) => 1+w*2 W: max window size (default 100) => 1+W*2
Bacterial genomes (few Mbp) use -max 100000 -W 10 Mammalian genomes (few Gbp) use -max 1000000 -W 100
- merge-anc.pl -max
max: max distance between the ref_intercept (default 1,000,000)
- Example:
cat PA14-PA7.delta | ~/bin/shrinkIds.pl | ~/bin/DELTA/delta2anc.pl > PA14-PA7.anc
cat PA14-PA7.anc | ~/bin/DELTA/filter-anc.pl -max 100000 -W 10 > PA14-PA7.filter.anc
PA14-PA7.filter.anc | ~/bin/DELTA/merge-anc.pl -max 100000 > PA14-PA7.merge.anc
cat PA14-PA7.merge.anc | ~/bin/DELTA/anc2delta.pl > PA14-PA7.merge.delta
wc -l *anc
691 PA14-PA7.anc
528 PA14-PA7.filter.anc
9 PA14-PA7.merge.anc
Pairwise Alignment Processing (poster)
Daniela Puiu*, Arthur Delcher, Steven Salzberg University of Maryland, College Park
Abstract
- We present a software tool for sequence alignment processing.
- The tool filters and clusters local alignments
- The tool was initially developed for finding syntenic regions in large eukariotic genomes but could also be used for analyzing rearrangements in microbial genomes.
- The software takes as input alignment data which is generally very fragmented (low identity) and noisy (repeats, rearrangements) and tries to filter and cluster it in
- A moving window approach is used for comparing each alignment to its neighbors and decide if it agrees or not with its neighbors. The alignments which clearly disagree get discarded.
- The process is run iteratively on increasingly larger window sizes till
Motivation
Sequence alignment is one of the most common bioinformatics problems. Due to its complexity there is no one general solution and multiple algorithms have been implemented for aligning different data set types. There is always a tradeoff between speed and accuracy.