Trinity is behaving like most normal unix program (at least when it comes to file paths), all files need to be specified by either an absolute path, a relative path or exist in the working directory.

If you want to specify multiple input files for --right and --left they need to be comma separated, and specified with a path that exists, you specified the path only to the first file argument and not to the other files, and therefore trinity does not find them. Assume the data is in /home/marziyeh/Data/RNA-seq\ data/Main\ data/ It might look simple, but the best way of doing this is:

cd /home/marziyeh/Data/RNA-seq\ data/Main\ data/;
/home/marziyeh/Software/trinityrnaseq-Trinity-v2.4.0/Trinity --seqType fq --max_memory 4G \ 
--left paired_32_L3_1_trimmed.fq.gz,paired_33_L3_1_trimmed.fq.gz  \
--right paired_32_L3_2_trimmed.fq.gz,paired_33_L3_2_trimmed.fq.gz --CPU 2  --output trinity_out_dir_32_32.fq.gz

Despite other comments, the space in the directory name is not the problem here because it is correctly quoted, even though it is a nuisance, less so with tab completion.

A few more hints for running trinity:

• Create a little shell-script that contains these commands, to be able to reproduce the parameters afterwards, also if any of the jobs fail during a run, you need to re-run the script on the exact command to finish the failed jobs (re-entrant)
• This process will run for a very very long time for any reasonably sized transcriptome, 4G is very little RAM and 2 CPUs too, try to provide as many CPUs and RAM as possible
• If Trinity is running on a server you definitely need to run it via nohup, otherwise the process will be interrupted by any logout, and the output will flood your terminal