I have exome sequencing data for oral rinse samples tested for hereditary cancers. I am looking for potential non-human (cow or pig) contaminants. What would be the best methodology to detect the contaminants? What so far what I have done is : 1. Align fastqs with human reference(bwa-mem) 2. Samtools to extract unmapped reads from aligned bam files 3. Build an index of unmapped reads assumed to be potential contaminants. [samtools view -u -f 12 -F 256] (both mates unmapped) 4. Map the unmapped reads with Cow and Pig reference (bwa-mem) 5. Extract mapped reads from this alignment 6. Confirm those which map exclusively to one ref. [Quality checks and coverages are calculated, reads with MQ 8 are considered for analysis]
Any suggestions are appreciated.
1 answer
FastQ Screen does exactly what you want:
FastQ Screen allows you to screen a library of sequences in FastQ format against a set of sequence databases so you can see if the composition of the library matches with what you expect.
Log in to answer this question.
-
Filtering reads that map to both references
written by michael.flower.14 •
-
Samtools mapped and unmapped flags, what is the difference between the flags?
written by DNAngel
-
Samtools SORT or bedtools bamToFastq giving error when converting reads to fastq.
written by DNAngel
-
Error when trying to convert BAM file to FASTQ.
written by DNAngel
-
Trying to understand samtools -f and -F
written by shibl_a •
-
difference between mapq=0 and samtools flag 0x900
written by bitpir
-
report only primary alignment BWA-MEM
written by Medhat
-
Find human contaminants in metagenomics microbiome studies
written by David
-
Read filtering strategy and BWA
written by Picasa
-
Explanation Of Low Mapping Quality!
written by khikho
An alternative could be bin the reads that map to pig/cow (or you could bin them to humans) using BBSplit from BBMap. This tool is designed for this specific application.