error running Merging contigs with minimus2

[DioufMouss0]

Hello I am trying to run my analysis with the merged mode as my 32G RAM are not enough for the co-assembly mode. The assembly goes well until the merging of the contigs with minimus2 as you can see below

STEP1 -> MERGING ASSEMBLIES: 01.merge_assemblies.pl Running cd-hit-est for removing redundant contigs Transforming to afg format Merging with minimus2 Error running command: /home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/bin/AMOS/minimus2 /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99 -D OVERLAP=100 -D MINID=95 -D THREADS=12 >> /home/abk/Desktop/EBFGSQM/syslog 2>&1 at /home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/scripts/01.merge_assemblies.pl line 71. Stopping in STEP1 -> 01.merge_assemblies.pl. Program finished abnormally Died at /home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/scripts/01.run_all_assemblies.pl line 233. Stopping in STEP1 -> 01.run_all_assemblies.pl. Program finished abnormally

Died at /home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/scripts/SqueezeMeta.pl line 941.

Here is the log file text

!!! 2024-02-04 23:21:41 Started by root@UNKNOWN on Sun Feb 4 23:21:41 2024

!!! 2024-02-04 23:21:41 Doing step 10: Building AMOS bank & Dumping reads !!! 2024-02-04 23:21:41 Running: rm -fr /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.bnk !!! 2024-02-04 23:21:41 Done! Elapsed time:0d 0h 0m 0s

!!! 2024-02-04 23:21:41 Doing step 11 !!! 2024-02-04 23:21:41 Running: /home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/bin/AMOS/bank-transact -c -z -b /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.bnk -m /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.afg START DATE: Sun Feb 4 23:21:41 2024 Bank is: /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.bnk 0% 100% AFG .................................................. Messages read: 9156570 Objects added: 9156570 Objects deleted: 0 Objects replaced: 0 END DATE: Sun Feb 4 23:22:52 2024 !!! 2024-02-04 23:22:52 Done! Elapsed time:0d 0h 1m 11s

!!! 2024-02-04 23:22:52 Doing step 12 !!! 2024-02-04 23:22:52 Running: /home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/bin/AMOS/dumpreads /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.bnk -M 0 > /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.ref.seq Objects seen: 4578284 Objects written: 4578284 !!! 2024-02-04 23:25:19 Done! Elapsed time:0d 0h 2m 27s

!!! 2024-02-04 23:25:19 Doing step 13 !!! 2024-02-04 23:25:19 Running: /home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/bin/AMOS/dumpreads /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.bnk -m 0 > /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.qry.seq Objects seen: 4578284 Objects written: 4578284 !!! 2024-02-04 23:27:43 Done! Elapsed time:0d 0h 2m 24s

!!! 2024-02-04 23:27:43 Doing step 20: Getting overlaps !!! 2024-02-04 23:27:43 Running: LD_LIBRARY_PATH=/home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/bin/AMOS/../../lib/mummer /home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/bin/AMOS/../mummer/nucmer --maxmatch --threads 12 -c 100 /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.ref.seq /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.qry.seq -p /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99 Killed !!! 2024-02-04 23:37:09 Command: LD_LIBRARY_PATH=/home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/bin/AMOS/../../lib/mummer /home/abk/Desktop/Array/Metagenomics/NGS_tools/SqueezeMeta/bin/AMOS/../mummer/nucmer --maxmatch --threads 12 -c 100 /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.ref.seq /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99.qry.seq -p /home/abk/Desktop/EBFGSQM/temp/mergedassemblies.EBFGSQM.99 exited with status: 137 !!! END - Elapsed time: 0d 0h 15m 28s

I don't have any idea of where the problem is could you please help with that?

[jtamames]

Hello

It looks like the system killed the process, probably because of lack of memory. You can try to analyze the data in sequential mode and combining later the results using combineSQM in SQMTools Best, J

[fpusan]

Yes, as discussed in the ReadMe

Note that the merged and seqmerge modes work well as a substitute of coassembly for running small datasets in computers with low memory (e.g. 16 Gb) but are very slow for analising large datasets (>10 samples) even in workstations with plenty of resources. Still, setting -contiglen to 1000 or higher can make seqmerge a viable strategy even in those cases. Otherwise, we recommend to use either the sequential or the co-assembly modes.

[DioufMouss0]

Yes, but I am having less than 10 samples (8) and 32GB RAM but I am still getting the error message

Aliou

---

From: Fernando Puente-Sánchez @.> Sent: Monday, February 5, 2024 11:30:50 AM To: jtamames/SqueezeMeta @.> Cc: Diouf Moussa @.>; Author @.> Subject: Re: [jtamames/SqueezeMeta] error running Merging contigs with minimus2 (Issue #789)

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.

Yes, as discussed in the ReadMe

Note that the merged and seqmerge modes work well as a substitute of coassembly for running small datasets in computers with low memory (e.g. 16 Gb) but are very slow for analising large datasets (>10 samples) even in workstations with plenty of resources. Still, setting -contiglen to 1000 or higher can make seqmerge a viable strategy even in those cases. Otherwise, we recommend to use either the sequential or the co-assembly modes.

— Reply to this email directly, view it on GitHubhttps://github.com/jtamames/SqueezeMeta/issues/789#issuecomment-1926668902, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AZ5SYCL3R65K6QFFL622WK3YSCYFVAVCNFSM6AAAAABCZGZ6SOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRWGY3DQOJQGI. You are receiving this because you authored the thread.Message ID: @.***>

[fpusan]

It still applies. The RAM requirements of minimus2 scale very fast (pretty sure it's quadratical) so it may not be working even if you had a big server (32 Gb is not much in this context).

[jtamames]

Citing 10 samples does not mean that you cannot run out of memory with less samples. We know that with more than 10 samples will probably happen, but can happen with less, for instance if they are big.

[fpusan]

Closing due to lack of activity, feel free to reopen