Hello

I think this question is for the package developers, but maybe somebody had the same problem. I am traying to analyze 59 bam files in a computer with 110GB of available memory usage, all bam files occupy 149GB. I can't obtain the splice graph and I don't get any error message.. I monitorized the process and it looks like the memory is full for this process, so maybe that is the reason. I parallelize the process with 12 CPU, but still it is not enough. To obtain the SGFeatureCounts class, all bam files are downloaded to R first? If it is so, I guess I would need to increase the memory.. Could you give me any advice how can I complete the process in my computer?

Thank you,

Magda

Hi Magda,

Can you please provide more details -- what commands were used to generate the SGFeatureCounts object and what package versions were used i.e. output of sessionInfo(). For larger analyses I strongly recommend not to use wrapper functions like analyzeFeatures() and instead run individual analysis steps separately (see section 12 of the vignette). This helps with troubleshooting and allows saving and inspecting of intermediate results. Please also see section 13 for recommendations on using multiple cores and typical memory requirements. 

Leonard