could not load Parasail module needed for Smith-Waterman

[border-info-nt]

Hii!!!

I want to use rmats-long from espresso.
I have the following problem.

I cannot run ESPRESSO_C.pl even though I have built Parasail.

Please help.

perl miniforge3/envs/espresso_env/bin/ESPRESSO_C.pl -I work_dir -F /Volumes/unnamed/GRCh38.primary_assembly.genome.fa -X 0 -T 2

Could not load Parasail module needed for Smith-Waterman

[EricKutschera]

It looks like you installed espresso from bioconda. In that case you shouldn't need to build the Parasail module

One possibility is that you installed the noarch bioconda package for espresso v1.5.0 which doesn't have the compiled Parasail module. From: https://anaconda.org/bioconda/espresso/files

There are two builds of v1.5.0:
noarch/espresso-1.5.0-hdfd78af_0.tar.bz2
linux-64/espresso-1.5.0-pl5321h43eeafb_1.tar.bz2

You can check which version you have with ls miniforge3/envs/espresso_env/conda-meta/espresso*. The linux-64 build includes lib/libparasail.so bin/Parasail.so bin/Parasail.pm

You can request conda to install that specific build with: conda install -c conda-forge -c bioconda espresso=1.5.0=pl5321h43eeafb_1

If you can't install that build then you can try downloading the source code from https://github.com/Xinglab/espresso and running ./src/Parasail/build

[border-info-nt]

Sorry.
I am new to bioinformatics, so please give me detailed instructions.

If I do it with -blast, will the results be different?

[EricKutschera]

If you run ESPRESSO_C.pl with --blast then you won't need the Parasail module. The results with and without --blast should be similar, but not identical. The main reason to avoid using --blast is that --blast is much slower

[brilliantlee2]

Thx，I'v solved this question.Firstly, install source package from https://anaconda.org/bioconda/espresso/1.5.0/download/linux-64/espresso-1.5.0-pl5321h43eeafb_1.tar.bz2.
Secondly, install conda package using conda install -c conda-forge -c bioconda espresso=1.5.0=pl5321h43eeafb_1.

After these steps, using bin/ESPRESSO_C.pl* in the espresso=1.5.0=pl5321h43eeafb_1 conda environment.