CUFFLINKS error: Warning: invalid start coordinate at line:

Question

[westin.kosater]

I downloaded a ChIP-seq dataset from ENCODE (ENCFF615OUG.bam) and made an indexed bam file of it (ENCFF615OUG.bai), then extracted the chromosome 22 sequence (chr22.bam) using samtools view

samtools view -b ENCFF615OUG.bam chr22 > chr22.bam

I need to get FPKM values for genes in chr22, so I tried running CUFFLINKS using the GTF file for chromosome 22 from UCSC table browser

cufflinks -G grch38-chr22.gtf chr22.bam

However when I run this command it outputs the following

[10:28:21] Loading reference annotation.
Warning: invalid start coordinate at line:
666 ENST00000615943.1   chr22   -   10736170    10736283    10736170    10736170    110736170,  10736283,   0   U2  none    none    -1,
Warning: invalid start coordinate at line:
668 ENST00000618365.1   chr22   -   10936022    10936161    10936022    10936022    110936022,  10936161,   0   CU459211.1  none    none    -1,
Warning: invalid start coordinate at line:
669 ENST00000623473.1   chr22   -   11065973    11067346    11065973    11067346    211065973,11067334, 11066015,11067346,  0   CU104787.1  incmpl  incmpl  0,0,
Error parsing strand (1) from GFF line:
669 ENST00000624155.1   chr22   +   11066500    11068089    11066500    11068089    211066500,11067984, 11066515,11068089,  0   BAGE5   cmpl    cmpl    0,0,

I'm not sure what causes these errors. I downloaded the gtf file right from UCSC Table so I'm not sure what to do to fix this error. Any ideas?

[Titus]

Can you show reads aligned from your bam ?

Answer 1

[westin.kosater]

I discovered my problem. When downloading from the UCSC table browser I did not select GTF file output

[Kevin Blighe]

Problem completely solved, then? If 'yes', I will move this to an answer.

[westin.kosater]

Yes, problem is solved