cuffmerge BAM file missing

[Parharn]

I am following the protocol by Trapnell, et al. On step 4 there is this command:
cuffmerge -g genes.gtf -s genome.fa -p 8 assemblies.txt

The txt file directs to transcripts.gtf so none of the files here are BAM! But I receive the following lines:

[bam_header_read] EOF marker is absent. The input is probably truncated.
[bam_header_read] invalid BAM binary header (this is not a BAM file).
File ./merged_asm/tmp/mergeSam_fileAqXvMn doesn't appear to be a valid BAM file, trying SAM...
[11:26:22] Loading reference annotation.
[11:26:23] Inspecting reads and determining fragment length distribution.
Processed 11334 loci.
> Map Properties:
> Normalized Map Mass: 22853.00
> Raw Map Mass: 22853.00
> Fragment Length Distribution: Truncated Gaussian (default)
> Default Mean: 200
> Default Std Dev: 80
[11:26:23] Assembling transcripts and estimating abundances.
Processed 11334 loci.


Why is it looking for BAM file? Then it looks that by trying SAM it can pass this step and give results, but should I be concerned about it?


Cheers,
Parham

[dpryan]

I recall that it does some sort of conversion of the GTF files to SAM. Yes, that seems a bit odd.

[Parharn]

Yes first thing it converts GTF to SAM. Do you have any suggestions? Here is the what I get before the BAM failure.

[Wed Jan 8 14:15:31 2014] Preparing output location ./merged_asm/
[Wed Jan 8 14:15:31 2014] Converting GTF files to SAM
[14:15:31] Loading reference annotation.
[14:15:31] Loading reference annotation.
[Wed Jan 8 14:15:32 2014] Quantitating transcripts
You are using Cufflinks v2.1.1, which is the most recent release.
Command line:
cufflinks -o ./merged_asm/ -F 0.05 -g genes.gtf -q --overhang-tolerance 200 --library-type=transfrags -A 0.0 --min-frags-per-transfrag 0 --no-5-extend -p 8 ./merged_asm/tmp/mergeSam_filen0cePq
[bam_header_read] EOF marker is absent. The input is probably truncated.
[bam_header_read] invalid BAM binary header (this is not a BAM file).

[Parharn]

Update:
I kept the temporary files while running cuffmerge. It builgs a SAM file from transcripts.gtf files. Then it seems first it reads it as BAM since I read samtools take the files as BAM default. Then it finds it is not a BAM so it tries SAM. Then it can read it and resume the task. I am writing the headers for the tmp SAM file here so you can see how it looks.
@HD VN:1.0 SO:coordinate
@SQ SN:2L LN: 22961179
@SQ SN:2LHet LN: 272340
@SQ SN:2R LN: 21142371
@SQ SN:2RHet LN: 3234607
@SQ SN:3L LN: 24536634
@SQ SN:3LHet LN: 2544176
@SQ SN:3R LN: 27894162
@SQ SN:3RHet LN: 2505998
@SQ SN:4 LN: 1275487
@SQ SN:U LN: 10047936
@SQ SN:X LN: 22417005
@SQ SN:XHet LN: 196035
@SQ SN:YHet LN: 340818
@SQ SN:dmel_mitochondrion_genome LN: 14916
@PG ID:cuffmerge VN:1.0.0

Please correct me if I got it wrong.

[dpryan]

That seems to fit what I recall. I vaguely remember that error message being pretty normal.

[yueluo]

Checking my previous cuffmerge runs, I got the same message every time but all the runs were successful. I guess we could just ignore it.