SqueezeMeta --restart in merged mode reruns 01.run_all_assemblies.pl instead of resuming merge step

[elahehseifollahi-hue]

Hello,

I am using SqueezeMeta v1.6.5 with 49 DNA samples and 49 RNA samples. Each sample has paired-end reads. My goal is to analyze the fungal species present in these samples and identify CAZymes.

Before running SqueezeMeta, I cleaned the reads with bbduk and fastp.

My first run used this command:

SqueezeMeta.pl
-m merged
-p MetaGenomeTranx
-s samples4.txt
-f "$INPUT/SqueezeMeta_input"
-a spades
-c 1000
-t 48
--extdb $Path/CAZy_DB/mydb.list
--nopfam

After about 14 days, the run stopped at:

STEP1 -> MERGING ASSEMBLIES: 01.merge_assemblies.pl

Then I tried to restart it with:

SqueezeMeta.pl
-m merged
-p MetaGenomeTranx
-s samples4.txt
-f "$INPUT/SqueezeMeta_input"
-a spades
-c 1000
-t 48
--extdb $Path/CAZy_DB/mydb.list
--nopfam
--restart

However, instead of resuming from the merge step, it started again from the beginning:

STEP1 -> RUNNING ASSEMBLY: 01.run_all_assemblies.pl (spades)

I would like to ask:

Why did --restart start again from the assembly step instead of resuming from 01.merge_assemblies.pl?
Is this the expected behavior in merged mode?
If a long run stops because of timeout, what is the correct way to continue from the last successful point?
Is there a way to avoid re-running STEP1 from the beginning and make the restart faster?

Thank you very much for your help.