Fail at jsm.py train

[GoogleCodeExporter]

What steps will reproduce the problem?
1.jsm.py train joint_snv_mix_two mm10_mouse_ref.fa 129tumor.bam 129normal.bam 
2.
3.

What is the expected output? What do you see instead?

[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).
Traceback (most recent call last):
  File "/users/jwu/bin/jsm.py", line 128, in <module>
    args.func(args)
  File "run.pyx", line 105, in joint_snv_mix.run.train (joint_snv_mix/run.c:2452)
  File "run.pyx", line 177, in joint_snv_mix.run.ModelFactory.get_counter (joint_snv_mix/run.c:3293)
  File "bam.pyx", line 19, in joint_snv_mix.samtools.bam.BamFile.__cinit__ (joint_snv_mix/samtools/bam.c:804)
  File "bam.pyx", line 42, in joint_snv_mix.samtools.bam.BamFile._open_file (joint_snv_mix/samtools/bam.c:1065)
Exception: File 
/aveo/data/reference/dna/mouse/mm10/prepared_by_bwa/mm10_mouse_ref.fa does not 
have valid header - is it SAM/BAM format?

What version of the product are you using? On what operating system?

JointSNVMix 0.8b2 on Ubuntu

Please provide any additional information below.

Original issue reported on code.google.com by zhenhuaw...@gmail.com on 7 Mar 2013 at 10:11

[GoogleCodeExporter]

It seems that JoingSNVMix expects a bam file where it is supposed to be a fasta 
file, not sure why?

Original comment by zhenhuaw...@gmail.com on 7 Mar 2013 at 10:12

[GoogleCodeExporter]

A few comments.

1) You appear to have the order of tumour and normal reversed. This won't cause 
the error but will be bad for results.

2) When you use `train` instead of `classify` there is a fourth mandatory 
argument which is the name of the file to write the learned parameters. This 
may fix the error. Try typing `jsm.py train -h` for the full list of commands.

Let me know if there is still an issue.

Cheers,
Andy

Original comment by AndrewJL...@gmail.com on 13 Mar 2013 at 12:45