These are chat archives for cboursnell/transfuse

Jul 2016
Jul 02 2016 01:23

I ran bamtools stats, but I am rather perplexed. When I ran transfuse on one pair of fastqs (R1 + R2), it ran to completion and created three bam files, one for each input assembly and one for the fused assembly. bamtools stats gives different numbers of "Total reads" for each of these bams, and none of them match the number of input reads:
Input reads (R1+R2): 64600962
Assembly 1 Total reads: 109765988
Assembly 2 Total reads: 206939154
fused assembly Total reads: 205490564

Does that make sense? Since the "Total reads" (above) per bamtools stats far exceeds my input reads, I will conclude that the alignment didn't finish for the "all samples" transfuse, and that the bam must be incomplete:

Input reads (all R1+R2): 1,645,830,924
Assembly total reads: 1,063,292,792

Is it possible to restart a transfuse run? Is there some way to debug why it "stalls"? And why are the bamtools stats reported reads so different from the actual input reads (for the one-sample run that completed)?