Error when combining metagenomes and metatranscriptomes with 'noassembly' 'nobinning' flags

[carleton-envbiotech]

Hello,

I am running SqueezeMeta v 1.7.2 (freshly installed) and using the following input code to look at a small number of test samples (included in the attached sample sheet), where I am trying to combine metagenomes (denoted by the DNA in the sample name) and metatranscriptomes (denoted by RNA in the sample name).

When using the following input code:

SqueezeMeta.pl \ -m sequential \ -s KB1_CF_DNA_RNA_sample_sheet_trial_short.txt \ -f /datastore/userdata/daniel/KB1_training_data/CF \ --a megahit \ -c 1500 \ -map bowtie \ -binners maxbin,metabat2,concoct \ -miniden 70 \ -extdb Custom_database_path_list_PFASDB_PlasticDB_RDaseDB_20260420.txt \ -t 20

I get the following error:

`Assembly not present in /datastore/userdata/daniel/KB1_training_data/CF/SQM/Iron_reduction_week2_rep1_RNA/results/01.Iron_reduction_week2_rep1_RNA.fasta, exiting
Stopping in STEP1 -> 01.run_all_assemblies.pl. Program finished abnormally

If you don't know what went wrong or want further advice, please look for similar issues in https://github.com/jtamames/SqueezeMeta/issues
Feel free to open a new issue if you don't find the answer there. Please add a brief description of the problem and upload the /datastore/userdata/daniel/KB1_training_data/CF/SQM/Iron_reduction_week2_rep1_RNA/syslog file (zip it first)
Died at /datastore/tools/gregoire/envs/SqueezeMeta-1.7.2/bin/SqueezeMeta.pl line 1060.`

I'll also note, that I was unable to add the --cleaning flag as well, Trimmomatic kept failing. Removing the flag allowed me to process the DNA samples, but then threw the error at the RNA samples.

Any help that could be provided on this small test data set would be greatly appreciated.

KB1_CF_DNA_RNA_sample_sheet_trial_short.txt

[fpusan]

The second column of your samples file contains a path rather than a file name.

DNA/Run2_CFcultureDNA_IronReduction_Week2_Bottle16_rep1_S10_R1_001.fastq.

which may be causing some errors (I suspect that's behind the problem with Trimmimatic in particular)

The recommended way would be to place your DNA and RNA fastq files into a single directory (you can put symlinks there, if you want to keep the real files separated) and pass that directory to SqueezeMeta with the -f flag

[fpusan]

Ah! But really the problem is much simpler.

You are running SqueezeMeta in sequential mode, which means that each sample is processed individually.
But for the RNA samples you have added the noassembly flag in the samples file, which keeps them from being assembled.
So for those samples there is no assembly present, and the pipeline fails.

The noassembly flag is meant to be used together with the coassembly, merged or seqmerge modes. In that cases, the metagenomes will contribute reads to the assembly even if the metatranscriptomes do not, so assembly will happen anyways.

[carleton-envbiotech]

Copying the files over to the same directory got Trimmomatic running again.

Is there a way to avoid having to move everything into the same directory? I'd like to keep my DNA and RNA files separate, as there are dozens of them and having them organized in dedicated subdirectories is helpful.

On a side note, are there any plans to incorporate potentially faster trimming steps, such as using fastp?

[fpusan]

You can keep the fastq files in their original directories. Just make a new one for the purpose of running SqueezeMeta and symlink everything there.

On fastp (or other cleanup software) I'll see how difficult It would be to add it. Generally, our policy has been that the user is responsible for providing clean reads.

[fpusan]

Can you share an example command with how you usually run fastp?

[carleton-envbiotech]

This is an example of an AI-assisted loop I generated to apply fastp, but I am in the midst of testing it:

`for r1 in DNA/*_R1_001.fastq.gz; do
r2=${r1/R1/R2}
base=$(basename "$r1" _R1_001.fastq.gz)

fastp \
-i "$r1" \
-I "$r2" \
-o "DNA_trimmed/${base}_R1_paired.fastq.gz" \
-O "DNA_trimmed/${base}_R2_paired.fastq.gz" \
--unpaired1 "DNA_trimmed/${base}_R1_unpaired.fastq.gz" \
--unpaired2 "DNA_trimmed/${base}_R2_unpaired.fastq.gz" \
--detect_adapter_for_pe \
--cut_front --cut_tail \
--cut_window_size 4 \
--cut_mean_quality 20 \
--length_required 50 \
--thread 20 \
--html "DNA_trimmed/${base}_fastp.html" \
--json "DNA_trimmed/${base}_fastp.json"

done`

[carleton-envbiotech]

I've had a separate issue arise after moving the files to the same directory and using Trimmomatic.

Using the following input code:

SqueezeMeta.pl \ -p KB1_coassembly_trial \ -m coassembly \ -s KB1_CF_DNA_RNA_sample_sheet_trial_short.txt \ -f /datastore/userdata/daniel/KB1_training_data/CF \ --cleaning \ --a megahit \ -c 1500 \ -map bowtie \ -binners maxbin,metabat2,concoct \ -miniden 70 \ -extdb Custom_database_path_list_PFASDB_PlasticDB_RDaseDB_20260420.txt \ -t 20

it looks like the process is crashing at the summary statistic step (error message output below):

[4 hours, 36 minutes, 24 seconds]: STEP21 -> MAKING FINAL STATISTICS: 21.stats.pl Illegal division by zero at /datastore/tools/gregoire/envs/SqueezeMeta-1.7.2/SqueezeMeta/scripts/21.stats.pl line 320. Stopping in STEP21 -> 21.stats.pl Died at /datastore/tools/gregoire/envs/SqueezeMeta-1.7.2/bin/SqueezeMeta.pl line 1003.

Trying to the use --restart and -step 21 flags provides the following and does not allow me to continue the run:

[0 seconds]: STEP22 -> MAKING SUMMARY TABLES: sqm2tables.py Traceback (most recent call last): File "/datastore/tools/gregoire/envs/SqueezeMeta-1.7.2/SqueezeMeta/utils/sqm2tables.py", line 235, in <module> main(parse_args()) File "/datastore/tools/gregoire/envs/SqueezeMeta-1.7.2/SqueezeMeta/utils/sqm2tables.py", line 63, in main total_reads, total_bases = parse_mappingstat(perlVars['$mappingstat']) File "/datastore/tools/gregoire/envs/SqueezeMeta-1.7.2/SqueezeMeta/utils/../lib/utils.py", line 64, in parse_mappingstat total_reads.append(int(line[1])) ValueError: invalid literal for int() with base 10: '' Stopping in STEP22 -> sqm2tables.py Died at /datastore/tools/gregoire/envs/SqueezeMeta-1.7.2/bin/SqueezeMeta.pl line 1022.