Can we use the same input fastqs for Travis test and our test datasets?

[gongyixiao]

These 4 .tsv files points to different BAM files. Maybe we can just use one set of fastq files and fwer .tsv files: https://github.com/mskcc/vaporware/blob/master/tests/test_make_bam_and_qc.tsv https://github.com/mskcc/vaporware/blob/master/test_inputs/lsf/test_make_bam_and_qc.tsv https://github.com/mskcc/vaporware/blob/master/test_inputs/aws/baby_files.tsv https://github.com/mskcc/vaporware/blob/master/test_inputs/aws/test_make_bam_and_qc.tsv

Also, I don't think these test .tsv files are necessary now. Maybe we can get rid of these: https://github.com/mskcc/vaporware/blob/master/test_inputs/aws/WGS_mark_test.tsv https://github.com/mskcc/vaporware/blob/master/test_inputs/aws/WES_breast25TN_test.tsv https://github.com/mskcc/vaporware/blob/master/test_inputs/lsf/WES_breast25TN_test.tsv https://github.com/mskcc/vaporware/blob/master/test_inputs/lsf/test_make_bam_and_qc.tsv https://github.com/mskcc/vaporware/blob/master/test_inputs/lsf/WGS_bam.txt

[evanbiederstedt]

Sure, I'm fine with this.

Maybe we can just use one set of fastq files and fwer .tsv files:

I'm guessing these were created to do testing with 2 vs. 4 pairs.

Also, I don't think these test .tsv files are necessary now. Maybe we can get rid of these:

I would create a branch reserved for an AWS testing sandbox. And then keep these files.

Personally, I don't see this as the biggest thing to worry about currently either. Things are not too messy, and we can polish up when we're done.

[gongyixiao]

Done in this pull request https://github.com/mskcc/vaporware/pull/204