search

JaneliaSciComp / msg

Multiplexed Shotgun Genotyping
http://genomics.princeton.edu/AndolfattoLab/MSG.html
11 stars 12 forks source link

feature: msgRun1.sh doesn't need to wait can be in parallel #6

Closed JaneliaSciComp closed 12 years ago

JaneliaSciComp commented 13 years ago 
       - msgRun1.sh doesn't need to wait can be in parallel
       - maps reads to parental genomes, must be after updating
       - msgRun2.sh needs run 0 and run 1
gregpinero commented 12 years ago

Hi guys.

I wanted to confirm, can msgRun1.sh be run at the very beginning of launch?

Right now we have it waiting for msgRun0-1.sh and msgRun0-2.sh to finish. Does it not need to wait for either of those?

Thanks,

Greg

gregpinero commented 12 years ago

Notes to self:

It seems like msgRun1 does need to wait until msgRun0-1 and 0-1 finish.

When the parent reads are provided, it is looking for the updated parent files. I'm hitting an error if they're not there.

It's ~line 152 of msg.pl that's looking for the updated parent files

David thinks this check for the parental genomes is not strictly necessary when using the cluster, because msg 2, which needs the genomes, will wait for msg 1. He thinks this is a line for when msg is run on a desktop.

I'll have to investigate further whether this check can be taken out when not on the cluster.