You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As discussed on Slack here the aim of this feature is to make the latest versions of STAR the default now for all routes through the pipeline. There seem to be a number of fixes in recent versions of STAR that overcome weird seg faults and other issues that have been reported by users. We have stuck to an older version until now because the pipeline by default uses older indices from AWS iGenomes. Ideally, we would be able to auto-detect when AWS iGenomes is being used and switch to the appropriate container.
The plan is to auto-detect whether a --fasta called genome.fa is being used as well as a --gtf called genes.gtf. We can then have a switch in the pipeline that uses this information to determine which version of the container to use. This is the native naming convention used by all references on AWS iGenomes. However, this won't work in instances where users have renamed custom references downstream to follow this convention. In these relatively rare instances I think it makes sense to then use a custom config to re-define the containers:
The option to build and push the latest STAR indices to AWS iGenomes was also discussed but this doesn't make sense with the impending shift to using Refgenie for reference genome management.
Description of feature
As discussed on Slack here the aim of this feature is to make the latest versions of STAR the default now for all routes through the pipeline. There seem to be a number of fixes in recent versions of STAR that overcome weird seg faults and other issues that have been reported by users. We have stuck to an older version until now because the pipeline by default uses older indices from AWS iGenomes. Ideally, we would be able to auto-detect when AWS iGenomes is being used and switch to the appropriate container.
The plan is to auto-detect whether a
--fasta
calledgenome.fa
is being used as well as a--gtf
calledgenes.gtf
. We can then have a switch in the pipeline that uses this information to determine which version of the container to use. This is the native naming convention used by all references on AWS iGenomes. However, this won't work in instances where users have renamed custom references downstream to follow this convention. In these relatively rare instances I think it makes sense to then use a custom config to re-define the containers:The text was updated successfully, but these errors were encountered: