Getting genome indices for specific ENSEMBL version

[zhuchcn]

We are re-analyzing an cohort of samples done several years ago, and we would like to use the same ENSEMBL/GENCODE version that was used previously. Is there any way to build the references with a specific ENSEMBL version? The version we are interested is ENSMBLE 100. I went across the documentation but did not see an easy solution. Although the '--ftp-ensembl-path' seems related?

[ndaniel]

Hi @zhuchcn

Sorry, the easy way would be to find someone which has those built files corresponding to some older version of FusionCatcher. It is not easy feature to backtrak Ensembl version in FusionCatcher because FusionCatcher integrates other 20+ databases. That means that very likely a parser that works for Ensembl release 100 does not work for Ensembl release 104 due to changes in Ensembl SQL database schema. The amount of work for doing that is big and unfortunatel here are not enough resources to support several Ensembl releases simultaneously.

[zhuchcn]

Thanks for the explanation! Just curious, if I use the fusioncatcher-build.py and specify the --ftp-ensembl-path to 'pub/release-100', is it not going to work? I haven't gotten it to work yet because our cluster don't have FTP access, so I'm currently finding a way to run it.

[duaghk]

@zhuchen I tried to use --ftp-ensembl-path=/pub/release-105 (I used gencode 39) and set --web=http://dec2021.archive.ensembl.org/info/data/ftp/index.html and it worked.