Extremely Long Search Time with FAIMS data

[Conor Jenkins]

Hello!

I was wondering if anyone else has experienced extremely long processing times of FAIMS data in PD. I compared the analysis time of MS Amanda to SEQUEST on a set of 60 samples run at 3 CVs each and SEQUEST processed in a couple of days on a computer with a new threadripper CPU while MS Amanda took upwards of a month. Does a newer version of MS Amanda handle FAIMS data better?

Best,

Conor Jenkins

[Viktoria Dorfer]

Hi Conor,

which version of PD and MS Amanda are you currently using? There are several ways how to influence the search speed of MS Amanda: in the Advanced Parameters there is a "6. Performance Settings" part. The higher the number you set here the faster it is going to be, but that of course strongly depends on the available RAM on your computer - the higher the numbers the more RAM needed. Can you try to change that?

In the configuration settings there is also a parameter "Percentage of available CPUs used" but that should be 100% by default, so that should be fine.

Best,

Viktoria

[Conor Jenkins]

Hello Viktoria,

Thank you for getting back to me. I have the following:

PD 2.4.15 and the most current MS Amanda from pd-nodes.org for PD 2.4 running with the below performance setting with 128 GB of RAM

Spectra per package: 50000

Protein Database Size: 8000000

It looks like the spectrum selector step is taking a very long time as well as MSAmanda. Watching the system information, it appears that there is only ~5% CPU utilization and I cant remember the memory utilization but it was never reaching above 50%

-C

[Viktoria Dorfer]

hmmm.. can you maybe provide me with one of your FAIMS data sets and a bug report of PD, such that I can have a look on what is really going on and taking so long?

Thanks,

Viktoria

[Conor Jenkins]

Here is a google drive link. 5 replicates per sample. HeLa Cells and the report from the last run (it crashed after about 3 weeks)

https://drive.google.com/drive/folders/1Uwv9c0i8Jww5JpqNMpF6h_cES4DOvWrc?usp=sharing

Thanks!

-C

[Viktoria Dorfer]

Hi Conor,

I wanted to give you an update on that! We are still investigating this! One parameter extending the search time is definitely the second search option you enabled, but it still should finish in a reasonable amount of time. Sorry to have no better answer on that yet, but I did not forget you!

Best regards,

Viktoria 

[Conor Jenkins]

Thank you very much for the update! I can provide more files for you to look at if you would like. Please let me know if there is anything that I can do to help

Best,

Conor

--
You received this message because you are subscribed to a topic in the Google Groups "MS Amanda" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/msamanda/9zGLVNhGc4w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to msamanda+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/msamanda/04a2938b-1efd-4e9e-9034-b5691edbe6c2n%40googlegroups.com.

[Conor Jenkins]

Hello Viktoria!

I was wondering if there was an update on this. Could it be the density of the MS/MS scans from the FAIMS acquisition?

Best,

Conor

[Viktoria Dorfer]

Hi Conor,

we just released a new version with some fixes concerning search time (v. 17952). I am not a 100% sure it is really fixing the FAIMS issue, although it worked for us, so could you maybe give it a try? It can be downloaded from the MS Amanda webpage: https://ms.imp.ac.at/?goto=msamanda 

Thanks for testing this and sorry for the delay!

Best regards,

Viktoria 

[Conor Jenkins]

Thank you very much! I will test and give you feedback as soon as possible. 

Best,

-C

To view this discussion on the web visit https://groups.google.com/d/msgid/msamanda/8dea907c-6db3-4198-a6f6-f8ff99087943n%40googlegroups.com.