Currently lost between contigs, scaffolds and chromosomes due to positionally sorted scRNAseq bam files
Hello everyone,
I have downloaded some scRNAseq bam files from EGA, but the contents of the files look like this (header + first 3 rows of a random bam file, accessed with samtools):
@HD VN:1.6 SO:coordinate
@SQ SN:AAACCTGAGAACAACT-1_contig_1 LN:685
@SQ SN:AAACCTGAGAACAACT-1_contig_2 LN:673
@SQ SN:AAACCTGAGAACAACT-1_contig_3 LN:478
There is clearly something wrong or missing here. So the question is, are these files analyzable at all, and if no, what info should I expect to see in the bam file of a scRNAseq alignment? Also, is CellRanger the only option for analyzing this kind of data? Thank you in advance for your time.
• 1,022 views
•
link
updated
by
ATpoint
•
written
by
nyxtoviopouli •
•
written
by
nyxtoviopouli •
0 answers
No answers yet.
Log in to answer this question.
More posts like this
-
Header too long for bam format
written by selplat21 •
-
Change chromosome notation in bam file
written by Sara •
-
Extract the SAM File header section
written by mhasa006
-
StringTie "missing header error" [for both HISAT2 and STAR alignment files]
written by EagleEye
-
extracting data from bam file
written by Sara
-
Tools for calculating TPM from BAM files
written by 309772486 •
-
cn-mops Error "Missing refSeqName"
written by karim •
-
BAM header file error
written by Bioblazer
-
What does the @HD GO record type mean in a BAM file?
written by zengtony743
-
samtools merge error
written by Hatem Elshazly •
I have never looked at scRNA-seq headers actually but if this is a CellRanger bam file then simply use the 10x Chromium bam2fastq utility to restore the original fastq files. From there you have full control over the preprocessing. CellRanger is one option. STARsolo, alevin-fry or kallisto-bustools are others that perform well.
That is an odd header. What kind of data is this? Can you link the EGA accession (if it is public)? Generally
cellrangershould produce BAM files aligned to the genome reference and files have a "normal" header. It looks like these are cellbarcodes attached to something.