Stopping in STEP4 —> 04.rundiamond .pl Program finished abnormally

[da-gunjan-arya]

All checks successfully after running test_install.pl When I tried to running with other sample in sequential mode for flye assembly it Stopping in STEP4 —> 04.rundiamond .pl Program finished abnormally at line 941

reading your previous issues #707 I let it auto detect the -b according to my RAM viz. (We will set Diamond block size to 15.2 [Gb RAM/8, Max 16] ). It was still producing the same error. This is the error "Unexpected end of input" detected in diamond.nr.log

and i'm attaching the following syslog file

syslog.zip

[fpusan]

This doesn't seem to be a problem with memory. Rather it would seem the problem happens when DIAMOND is trying to load the reference database. Did you get the database using download_databases.pl?

[fpusan]

Also, did the pipeline work with the test datasets?

[da-gunjan-arya]

Database was re-downloaded using download_databases.pl and yes it works for test dataset. The script is able to load reference db and perform diamond, but now it stops at step 10. viz. bowtie2-align (in this it consumes up all the RAM)

[fpusan]

Please help me confirm that I understood your issue correctly. So after re-downloading the database: 1) SqueezeMeta works for the test dataset and the pipeline completes fully. 2) With your own dataset, step 4 now also works correctly. 3) However, with your own dataset, the pipeline now stops at step 10 in which bowtie2 is consuming all the RAM. Are these three points correct?

[da-gunjan-arya]

Yes, all points are correct.

[fpusan]

Can you manually run the command producing the error (shown after the "An error occurred during mapping with command:" message) removing the --quiet flag? Let's see if we get a more specific error message

[fpusan]

How much RAM do you have available?

[da-gunjan-arya]

I will try running that. Available RAM 124gb

[fpusan]

I see that you used flye as an assembler. Are these long reads? If so you should be using minimap2 for mapping. Check the ReadMe for instructions on how to do this.

[da-gunjan-arya]

I've tried running it with -map minimap2-ont and it was sorted but since sequencing data had more short reads even though its nanopore, I preferred using default parameters.

[jtamames]

Do you mean that minimap2-ont is working for you? It does not matter the length of the reads, minion reads have particular errors that makes it very advisable the usage of minimap2-ont

[da-gunjan-arya]

Yes it's working for now with all the samples I tried with. The pipepline runs without any errors. Thanks for the help.

[fpusan]

Glad to hear! closing issue