Taxor Profiling Module Not Utilizing Multiple Cores

[tthye1]

Dear Jens,

Thank you for this excellent tool. I noticed that when running the profile module, taxor only utilizes one core, even though 20 cores have been assigned. As a result, the process has been running now for two days. In contrast, the search module effectively uses multiple cores.

Best,

Thorsten

Applied code:

taxor profile
--threads 20
--search-file /data/out_META_00_MAG/C10342-TZ_NP.search.txt
--cami-report-file C10342-TZ_NP.report
--seq-abundance-file C10342-TZ_NP.abundance
--binning-file C10342-TZ_NP.binning
--sample-id C10342-TZ_NP

[JensUweUlrich]

Dear Thorsten,

thanks fo your interest in Taxor. I just recently did some bug fixes and performance improvements in Taxor, which are implemented and provided in the latest version 0.2.0. The version is actually not merged into the bioconda main branch, but I hope this will happen within the next days.
You are absolutely right that the profiling module is not multi-threaded. I tried to implement multi-threading for several code blocks, but without any performance improvements. Since the performance in the latest version is satisfying, I removed the thread option and thus offered a parameter to provide the number of EM steps.
Please do also not use the actual pebuilt GTDB index file, which will give you wrong results. I will also upload a fixed index file within the next days. I'll let you know when both the new index and the new bioconda version of taxor are ready to use.

[tthye1]

Dear Jens Uwe, Thank you for the quick update. I'm delighted to hear that the GTDB index file appears to be faulty, as I encountered some unusual results in my recent analysis. This saves me from additional error-checking ;) Thanks and best wishes, Thorsten

…

On Feb 28, 2025, at 10:52, Jens-Uwe Ulrich ***@***.***> wrote: JensUweUlrich left a comment (JensUweUlrich/Taxor#15) Dear Thorsten, thanks fo your interest in Taxor. I just recently did some bug fixes and performance improvements in Taxor, which are implemented and provided in the latest version 0.2.0. The version is actually not merged into the bioconda main branch, but I hope this will happen within the next days. You are absolutely right that the profiling module is not multi-threaded. I tried to implement multi-threading for several code blocks, but without any performance improvements. Since the performance in the latest version is satisfying, I removed the thread option and thus offered a parameter to provide the number of EM steps. Please do also not use the actual pebuilt GTDB index file, which will give you wrong results. I will also upload a fixed index file within the next days. I'll let you know when both the new index and the new bioconda version of taxor are ready to use. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread. <#15 (comment)> <https://github.com/notifications/unsubscribe-auth/AHWFHMVJODGU3VCYDT3JCMD2SAWUDAVCNFSM6AAAAABX5XBBMKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMOJQGIYDQMRVGA> JensUweUlrich left a comment (JensUweUlrich/Taxor#15) <#15 (comment)> Dear Thorsten, thanks fo your interest in Taxor. I just recently did some bug fixes and performance improvements in Taxor, which are implemented and provided in the latest version 0.2.0. The version is actually not merged into the bioconda main branch, but I hope this will happen within the next days. You are absolutely right that the profiling module is not multi-threaded. I tried to implement multi-threading for several code blocks, but without any performance improvements. Since the performance in the latest version is satisfying, I removed the thread option and thus offered a parameter to provide the number of EM steps. Please do also not use the actual pebuilt GTDB index file, which will give you wrong results. I will also upload a fixed index file within the next days. I'll let you know when both the new index and the new bioconda version of taxor are ready to use. — Reply to this email directly, view it on GitHub <#15 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AHWFHMVJODGU3VCYDT3JCMD2SAWUDAVCNFSM6AAAAABX5XBBMKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMOJQGIYDQMRVGA>. You are receiving this because you authored the thread.