[miso-users] Creating custom indexed gff files

[Sahin Naqvi]

Hi Yarden,

I am trying to make my own custom gff file of X and Y-linked genes to estimate exon PSI values with MISO, and am having some problems. I have written my exons of interest to file as such:

chrX	SE	exon	44879855	44879975	.	+	.	ID=KDM6A	

I'm fairly sure this is the write pattern, and it does index alright, but run_events_analysis crashes:

WARNING: It looks like your GFF annotation file and your BAM file might not have matching headers (chromosome names.) If this is the case, your run will fail as no reads from the BAM could be matched up with your annotation.

Please see:
	http://genes.mit.edu/burgelab/miso/docs/
 for more information.
It looks like your BAM file (/lab/solexa_page/snaqvi/tophat/ERR030886/sorted/accepted_hits.sorted.bam) contains 'chr' chromosomes (UCSC-style) while your GFF annotation (/lab/page/snaqvi/annotations/hg19/SE_xydeg_indexed) does not.

The first few BAM chromosomes were: 
chr1
BAM references: 
('chr1', 'chr10', 'chr11', 'chr12', 'chr13', 'chr14', 'chr15', 'chr16', 'chr17', 'chr18', 'chr19', 'chr2', 'chr20', 'chr21', 'chr22', 'chr3', 'chr4', 'chr5', 'chr6', 'chr7', 'chr8', 'chr9', 'chrM', 'chrX', 'chrY')
The first few GFF chromosomes were: 


The one thing I am doing different in my custom gff file is that I have no 'gene' or 'mRNA' entries, do I need these in order to run it correctly? I thought I wouldn't because each exon is treated independently…

Thanks,
Sahin