mikeaxtell / butter Goto Github PK

Realized that samtools depth has an internal max of 8000. This is a problem for butter because, internally when measuring bin coverages, it uses samtools depth. These comparisons will be inaccurate for an any bins where a depth exceeds 8000. Will need to re-write that sub-routine to avoid use of samtools depth.

I got an issue at the third step of bam2wig, and the program failed to open the temp file "chr_lens_temp.txt".

I have installed wigtobigwig. My BAM file has been sorted and indexed. The header is as follows,

$ samtools view -H PB_ATAC_T2.filtered.bam
@hd VN:1.5 SO:coordinate
@sq SN:1 LN:307041717
@sq SN:2 LN:244442276
@sq SN:3 LN:235667834
@sq SN:4 LN:246994605
@sq SN:5 LN:223902240
@sq SN:6 LN:174033170
@sq SN:7 LN:182381542
@sq SN:8 LN:181122637
@sq SN:9 LN:159769782
@sq SN:10 LN:150982314
@sq SN:Pt LN:140384
@sq SN:Mt LN:569630
....

Why did the error occur? How to fix it?

Thanks.

In specialized alignment scenario (to a virus genome) where there are no higher-level multi-mappers, version 0.2.4.1 crashes. It should just complete the alignment. Will fix in next release.

When run with option --no_condense, the progress tracker during bowtie alignment states "x million non-redundant reads aligned and counting" .. under --no_condense, it should just be regular reads, not non-redundant reads.

Hi Mike,

Im running butter using:

Ubuntu 16.04.1 LTS
kernel 4.4.0-38-generic
bowtie version 1.1.2
samtools 1.3.1

However, after bowtie mapping it crashes.

butter foo_1_trimmed.fq bar.fa

butter version 0.3.3
Fr 7. Okt 16:07:08 CEST 2016
Host:
Working Directory:
Genome: bar.fa
Reads: foo_1_trimmed.fq
Reads format: q
Adapter: None. Reads are assumed to be trimmed already.
Read condensation: ON
Max Mismatches: 0
Max number of possible placments for any read: 1000
Max number of possible placement for reads not guideable by density: 3
Processor cores for bowtie: 1
bam2wig setting: combined
Checking dependencies
samtools: /usr/bin/samtools
bowtie: /usr/bin/bowtie
bam2wig: /home/foo/tools/butter/bam2wig
wigToBigWig: /home/foo/bin/wigToBigWig
bowtie index files for genome bar.fa: Found
Compressing trimmed reads to a set of non-redundant queries ..
Reads in: 3047595
Non-redundant reads out: 2085386
File: foo_1_trimmed_condensed.fasta
Bowtie iteration one: 2085386 queries representing 3047595 reads
Calling bowtie with command:
bowtie -f -v 0 --best --strata --all -m 1 --max foo_trimmed_condensed_mmap1.fasta -p 1 -S bar.fa foo_1_trimmed_condensed.fasta
Progress (dots indicate one percent of queries):

After that I get prompt again. File condensed.fasta is populated but .condensed_mmap1.fasta is not.

Thanks in advance!

Best Regards

AndreiR

I have these in the same folder as the path for the genome fa = hg19.fa
However, this does not detect the ebwt files and constantly builds the index using bowtie build.
Is there anything wrong in the current index that it does not detect them?

hg19.1.ebwt hg19.2.ebwt hg19.3.ebwt hg19.4.ebwt hg19.fa hg19.rev.1.ebwt hg19.rev.2.ebwt make_hg19.sh

Thanks,

User reports fatal exit when using butter in the mapping phase of a ShortStack run. Looks like an unexpected bowtie output in the second iteration, where none of the query reads should be unmapped so none should have XM:i or 0. Only happens with some of user's files, not all. Will investigate...

Suggestion to have option to create a wiggle genome-browser file from the completed BAM file. Probably conversion to bigWig best left to user, since I don't want to make the UCSC script wigToBigWig a dependency. Thanks to Julie Law for this suggestion.