The main issues I keep encountering are as follows...
# denovotranscript
t
The main issues I keep encountering are as follows: 1. Making a master assembly from Trinity, SC- and RNA-SPADES, involves combining the outputs of these processes for each assembly input channel. Using
.mix
for this is inappropriate because this process doesn't wait for all the assemblies to finish.
.collect
will combine all the different assembly inputs together instead of just the one channel input. I circumvented this with
.collectFile
which writes a file for each based on the
meta.id
identifier to distinguish the input assemblies, and then I read the file back in as a channel. The issue with this method is that it isn't neat and it will wait for every single assembly input to be finished before moving on, even though in theory each assembly could be processed the moment it is done. This is complex to explain so I'd be happy to try and expand upon this if anyone wants me to. 2. Memory issues are a big problem especially for Sortmerna, HISAT and Salmon. The
cat_later
branch helps with this since the inputs for the processes are smaller, but for large input files the pipeline regularly crashes. 3. Caching is not working properly but I have not tried to fix this at all yet.