Hi, I try to run the command nextflow run nf-core/rnasplice --input samplesheet.csv --source fastq --contrasts contrastsheet.csv --outdir /home/fcastillo/proyecto_p1/parte1/results_RNAsplice3 --aligner star_salmon --save_align_intermeds --genome GRCh37 -profile docker. (My data is a paired end https://www.ncbi.nlm.nih.gov/Traces/study/?acc=SRP277841&o=acc_s%3Aa). But I'm getting an error when trying to run, I wanted to know if anyone has had a similar error and how you solved it please so you can explain it to me. Trate de ejecutarlo tambien con otro comando pero nuevamente tengo el mismo error nextflow run nf-core/rnasplice --input samplesheet.csv --source fastq --contrasts contrastsheet.csv --outdir /home/fcastillo/proyecto_p1/parte1/results_RNAsplice4 --pseudo_aligner salmon --suppa --genome GRCh37 -profile docker Sorry for the inconvenience again.

The process was killed, because it requested 72 GB of RAM, and only 62 GB were available. While it is possible that a process may need more than 62 GB memory, this is not necessarily always the case, and your error is primarily a problem with the configuration. Where are you running this, a stand-alone server? If you don't have multiple nodes,

-profile docker

might be sufficient for most of the configuration settings, but you will still have to tell the pipeline the limits of your machine by specifying the parameters

--max_memory

(probably

62.GB

in your case,

--max_time

(this is mostly relevant for systems with schedulers, which kill processes that go over time), and

--max_cpus

(number of CPU cores in your system; you could subtract one or two for the Nextflow process itself, especially if have a lot). If you don't know the number of CPU cores, run this:

cpus=$(grep '^physical id' /proc/cpuinfo | sort -u | wc -l)
corespercpu=$(grep -m 1 '^cpu cores' /proc/cpuinfo | grep -oP '\d+$')
echo $((cpus * corespercpu))

If you have a system with multiple nodes and a scheduler, you may want to add additional configuration to take advantage of it (the default local executor only runs on the machine where you start the pipeline). If a process needs more memory than your machine can provide, you will have to either find a way through the pipeline that does not use that process (depending on what it is, there may not be one) or find a different system to run to pipeline. ¡Buena suerte!