I started a local galaxy on a local server. I was able to insert the local reference genomes for mapping in Bowtie2 and BWA-MEM along with doing indexes for it. Mapping works great and without problems.
I wanted to do variant calling for created BAM files (FreeBayes or VarScan) and I have a problem because this local reference genome that I used for mapping does not appear in these tools. If I make a variant calling on the file hg19.fasta from history, then the numbering unfortunately does not match.
How to make these reference genomes (from Bowtie2 or BWA) available in other tools?
So I assume you’ve been using the corresponding data_managers to build the Bowtie2 and BWA-MEM indices?
What freebayes and varscan need is a genome and its samtools .fai index. There is a dedicated data manager for this type of index, too, and it’s called sam_fasta_index_builder. Run that on your ref genomes and they should appear in freebayes, varscan, and quite some other tools.
Best,
Wolfgang
Hi @tomlodz – just to add a bit to @wm75’s reply:
To avoid problems, index reference genomes using the four primary Data Managers first (the sam index is included). This ensures basic functionality across tools, and Galaxy’s built-in functions, that do not have a tool-specific Data Manager.
More help in the post below. It lists the four primary Data Managers that should be run on every genome and the order that tends to work best:
These four can be run in a different order or after other tool-specific DMs are run, but that sometimes leads to problems.
For this genome, run the sam index, plus Picard and the twoBit (the “fasta fetch” genome DM was already run when you indexed for the mapping tools).
When you index your next genome, run those four DMs first, in the recommended order, then any tool-specific DMs.
To review how others have solved various problems when creating indexes in prior Q&A, click on the “data-managers” tag I added to your post.