search

privefl / bigsnpr

R package for the analysis of massive SNP arrays.
https://privefl.github.io/bigsnpr/
186 stars 44 forks source link

Computational time and memory #87

Closed MehreenSoomro closed 4 years ago

MehreenSoomro commented 4 years ago

Hi, I have been trying to use LDpred2 without any success. Every time I try to calculate snp_cor either R crashes or give error like cannot allocate memory. In other instance, I tried to generate svd plot for mean and standard deviation, it is running since two days without any outcome. I have dataset from UK biobank for Psa with 1200 individuals, using 4 cores on 500000 variants. Also, as mentioned in the tutorial the analysis is parallelized per chromosome, so far I have not been able to get results out of first chromosome. Initially, I was using university cluster icsf which had RAM of 256GB allocated to it and after the memory issues, I shifted my analysis to csf3 but receiving similar problems. I will be extremely grateful, if you can help in pointing out any solution to this problem or is there something wrong in my analysis.

Best Regards, Mehreen

privefl commented 4 years ago

I can help you but I will need more information.

Let's start with:

  1. Which versions (bigstatsr/bigsnpr) are you using?
  2. Are you able to run the LDpred2 vignette using the data provided there?
MehreenSoomro commented 4 years ago

Thanks for your timely reply...

I am using bigsnpr 1.4.0...everything else runs smoothly all the background checks...structure of data...until it comes to the point where snp_cor is used and it shows memory issue there. I dont think I can use data provided in vignettes on my university server. I was unable to install bigsnpr version 1.4.0 on my home PC due memory issue again...[Image] [Image]

Sent from Outlook Mobilehttps://aka.ms/blhgte

privefl commented 4 years ago

Memory issue when installing the package?!

Why wouldn't you be able to download the data there?

MehreenSoomro commented 4 years ago

It's a secure environment...nothing external can be stored in those directories apart from the university provided data...

Sent from Outlook Mobilehttps://aka.ms/blhgte

From: Florian Privé notifications@github.com Sent: Sunday, May 31, 2020 1:48:58 PM To: privefl/bigsnpr bigsnpr@noreply.github.com Cc: MehreenSoomro mahatyab@hotmail.com; Author author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

Memory issue when installing the package?!

Why wouldn't you be able to download the data there?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-636466996&data=02%7C01%7C%7C77590dfe184e4b59166e08d80560fcb2%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637265261394302218&sdata=69%2Fr1uXX0GGJPl6vpgO0%2F5RUzdtP5kzEfpIoAdod7%2FQ%3D&reserved=0, or unsubscribehttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYRTLOZPEUIJTE5BWOTRUJG3VANCNFSM4NPENQJQ&data=02%7C01%7C%7C77590dfe184e4b59166e08d80560fcb2%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637265261394312211&sdata=YSHj6EwJBO82V4LztCTDfDRgval2Nsn30NVIhHwe3FI%3D&reserved=0.

privefl commented 4 years ago

I recommend that you ask to get the data there and try with this data. How were you able to install the package if you can't get things into the cluster?

MehreenSoomro commented 4 years ago

Ok I will try it with this data and let you know if it shows the similar issue on it as well. Thanks a lot.

Sent from Mailhttps://go.microsoft.com/fwlink/?LinkId=550986 for Windows 10

From: Florian Privémailto:notifications@github.com Sent: 31 May 2020 14:09 To: privefl/bigsnprmailto:bigsnpr@noreply.github.com Cc: MehreenSoomromailto:mahatyab@hotmail.com; Authormailto:author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

I recommend that you ask to get the data there and try with this data. How were you able to install the package if you can't get things in the cluster?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-636469408&data=02%7C01%7C%7C72d8a57b461d4589fc8808d80563eb20%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637265273982944058&sdata=xzxHSR1nD%2FzuQkbrNUnK%2FtIXpKDgqn22j5r8xZFgGJ4%3D&reserved=0, or unsubscribehttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYROV35KZIKC4JATV2LRUJJKLANCNFSM4NPENQJQ&data=02%7C01%7C%7C72d8a57b461d4589fc8808d80563eb20%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637265273982954054&sdata=ifn5VPeqU1zuHW6qRTeZy%2BRAa%2Fo8rx7hCU9YHrMD%2Bh0%3D&reserved=0.

MehreenSoomro commented 4 years ago

Still the same issue.

Best Regards, Mehreen

Sent from Outlook Mobilehttps://aka.ms/blhgte

From: Florian Privé notifications@github.com Sent: Sunday, 31 May 2020, 13:23 To: privefl/bigsnpr Cc: MehreenSoomro; Author Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

I can help you but I will need more information.

Let's start with:

  1. Which versions (bigstatsr/bigsnpr) are you using?
  2. Are you able to run the LDpred2 vignette using the data provided there?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-636464068&data=02%7C01%7C%7C8db80950f1634cea6d3308d8055d67fd%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637265246012788688&sdata=50xsMVxGe7LRkewMpbIVWk%2Fwui6qFTVuzzs6Ewgbfa8%3D&reserved=0, or unsubscribehttps://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYW2KYYWWTD34NYCUMLRUJD3RANCNFSM4NPENQJQ&data=02%7C01%7C%7C8db80950f1634cea6d3308d8055d67fd%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637265246012788688&sdata=DQ94yoENBUs1Hy2slILYP6Shiv6MTErKS1tY9dISKkM%3D&reserved=0.

privefl commented 4 years ago

This can't be a memory issue.. How do you install the package? Are you using conda? Can you try a clean install?

MehreenSoomro commented 4 years ago

Hi Florian, Thanks for your reply. Today, I managed to run the algorithm on public data and worked absolutely fine. I will be extremely grateful, if you can guide me through what is missing from my dataset which is causing the memory issue.

Best Regards, Mehreen

Sent from Mailhttps://go.microsoft.com/fwlink/?LinkId=550986 for Windows 10

From: Florian Privémailto:notifications@github.com Sent: 31 May 2020 21:06 To: privefl/bigsnprmailto:bigsnpr@noreply.github.com Cc: MehreenSoomromailto:mahatyab@hotmail.com; Authormailto:author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

This can't be a memory issue.. How do you install the package? Are you using conda? Can you try a clean install?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-636521765&data=02%7C01%7C%7C59fc36ebedfd439a251f08d8059e1415%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637265523777774755&sdata=KhEDq62Y3JObqn2PJyYAdzGdyxCAlvJYOcZRBe%2Bg0q0%3D&reserved=0, or unsubscribehttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYWV6IYLZV2RWLWRU43RUK2DRANCNFSM4NPENQJQ&data=02%7C01%7C%7C59fc36ebedfd439a251f08d8059e1415%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637265523777784749&sdata=%2B9T9rfIiwS01S8sGzCc9BWvnl8%2BHph7KdNIbn3lCwdc%3D&reserved=0.

MehreenSoomro commented 4 years ago

Sent from Mailhttps://go.microsoft.com/fwlink/?LinkId=550986 for Windows 10

From: mehreen soomro mahatyab@hotmail.com Sent: Wednesday, June 3, 2020 12:29:47 PM To: mehreen soomro mahatyab@hotmail.com Subject: RE: [privefl/bigsnpr] Computational time and memory (#87)

Sorry to bother you a lot, but I am also receiving this error.

Sent from Mailhttps://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgo.microsoft.com%2Ffwlink%2F%3FLinkId%3D550986&data=02%7C01%7C%7Cefca7b66e5c44038f6d408d807b16d8f%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637267805935826988&sdata=RMa%2FjS4glUSN%2FOeKNq3rRmbqnfZhkTDhuK7N38l%2B65w%3D&reserved=0 for Windows 10

From: mehreen soomromailto:mahatyab@hotmail.com Sent: 03 June 2020 12:16 To: privefl/bigsnprmailto:reply@reply.github.com Subject: RE: [privefl/bigsnpr] Computational time and memory (#87)

Hi Florian,

Thanks for your reply. Today, I managed to run the algorithm on public data and worked absolutely fine. I will be extremely grateful, if you can guide me through what is missing from my dataset which is causing the memory issue.

Best Regards,

Mehreen

Sent from Mailhttps://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgo.microsoft.com%2Ffwlink%2F%3FLinkId%3D550986&data=02%7C01%7C%7Cefca7b66e5c44038f6d408d807b16d8f%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637267805935837033&sdata=29Tr0qjJWzR%2FX01ns%2BUkW7dODQUCWhGApQWcIner%2B10%3D&reserved=0 for Windows 10

From: Florian Privémailto:notifications@github.com Sent: 31 May 2020 21:06 To: privefl/bigsnprmailto:bigsnpr@noreply.github.com Cc: MehreenSoomromailto:mahatyab@hotmail.com; Authormailto:author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

This can't be a memory issue.. How do you install the package? Are you using conda? Can you try a clean install?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-636521765&data=02%7C01%7C%7Cefca7b66e5c44038f6d408d807b16d8f%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637267805935847024&sdata=Vk7Zc96gpYjqnCs7QpAbnM%2FOF70zcPBcWgr5B6x49%2FE%3D&reserved=0, or unsubscribehttps://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYWV6IYLZV2RWLWRU43RUK2DRANCNFSM4NPENQJQ&data=02%7C01%7C%7Cefca7b66e5c44038f6d408d807b16d8f%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637267805935856989&sdata=PpMjvLoYV3z42LXWD4UiMcLJ%2BWJRk8ADLidXpikzxAQ%3D&reserved=0.

asasasasas

privefl commented 4 years ago

Hard to help you without more information.

Have you tried running snp_cor() on chromosome 22? How many variants is this corresponding to?

privefl commented 4 years ago

Any update on this?

MehreenSoomro commented 4 years ago

Hi Florian,

I hope you are doing well. Yes, we were able to run Ldpred2 algorithm on all the chromosomes after reducing the number of SNPs from 400,000 to 20,000 or less. Performance is poor overall and heratability for some of the chromosomes is greater than 1. Also couldnt sum the result from all 22 chromosomes.

Best Regards, Mehreen

Get Outlook for Androidhttps://aka.ms/ghei36

From: Florian Privé notifications@github.com Sent: Monday, July 6, 2020 9:40:49 PM To: privefl/bigsnpr bigsnpr@noreply.github.com Cc: MehreenSoomro mahatyab@hotmail.com; Author author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

Any update on this?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-654454278&data=02%7C01%7C%7Cf380864b26734d62509308d821ecde39%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637296648501904836&sdata=eK%2F0%2FygIjdtA%2Bhr%2FKP3ErHhXIWYo1oueqvkASqwHjNQ%3D&reserved=0, or unsubscribehttps://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYW6CJRCTXHIPNWYCKDR2IZFDANCNFSM4NPENQJQ&data=02%7C01%7C%7Cf380864b26734d62509308d821ecde39%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637296648501904836&sdata=PDv6wYxplLnD3ze%2BhawXLmOi3CCabUsPfnnIEY2MUGU%3D&reserved=0.

privefl commented 4 years ago

What are the sumstats? Linear, logistic? Have you checked the supp note on quality control of sumstats?

MehreenSoomro commented 4 years ago

Hi Florian,

Yes, we have checked the quality control note provided and followed all the instructions. But as soon as the number of variants reduce the algorithm was working fine. With higher number of variants it keeps on crashing or showing issue like cannot allocate memory. Summary statistics are logistic regression.

Best Regards, Mehreen

Get Outlook for Androidhttps://aka.ms/ghei36

From: Florian Privé notifications@github.com Sent: Tuesday, July 7, 2020 8:44:55 AM To: privefl/bigsnpr bigsnpr@noreply.github.com Cc: MehreenSoomro mahatyab@hotmail.com; Author author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

What are the sumstats? Linear, logistic? Have you checked the supp note on qualifying control of sumstats?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://eur06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-654663590&data=02%7C01%7C%7C6317c8219aeb4eacf86708d82249a4c3%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637297046968432227&sdata=tybns%2B4J2iux%2FWYpyFnQCquv7r7QK7H7qEAFDcTl6Jk%3D&reserved=0, or unsubscribehttps://eur06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYWB4T62JTRGVXSFH4DR2LG7PANCNFSM4NPENQJQ&data=02%7C01%7C%7C6317c8219aeb4eacf86708d82249a4c3%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637297046968432227&sdata=4qg31%2BQMuku%2BndPZKRVEe9wOtnYCoPaovNymC7VDXy4%3D&reserved=0.

privefl commented 4 years ago

What do you mean by performance is poor then?

And that you cannot sum results from all 22 chromosomes?

MehreenSoomro commented 4 years ago

I meant I have results from all 22 Chromosomes each chromosome separately and mean AUC from each is not more than 55 but I havent sum the results to get overall genome-wide performance. As I am still not sure how to correctly sum the results.

Best Regards, Mehreen

Get Outlook for Androidhttps://aka.ms/ghei36

From: Florian Privé notifications@github.com Sent: Tuesday, July 7, 2020 2:02:46 PM To: privefl/bigsnpr bigsnpr@noreply.github.com Cc: MehreenSoomro mahatyab@hotmail.com; Author author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

What do you mean by performance is poor then?

And that you cannot sum results from all 22 chromosomes?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://apc01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-654841434&data=02%7C01%7C%7C0c458104e16849fce6f208d822760cee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637297237697473886&sdata=tp%2FTS4kNvLGww88t%2B0M07NJxxe9byrDg7G5m9Z4oy3U%3D&reserved=0, or unsubscribehttps://apc01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYTNCRHLUXW5VET73O3R2MMHNANCNFSM4NPENQJQ&data=02%7C01%7C%7C0c458104e16849fce6f208d822760cee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637297237697483881&sdata=qBNR2QACWmLkpni4FHRm%2FupO%2F58dB5vcR5sy0YZe1Pg%3D&reserved=0.

privefl commented 4 years ago

It should be fine to just sum the predictions from all 22 chromosomes.

MehreenSoomro commented 4 years ago

Hi Florian,

I would be extremely grateful if you can provide an example code from any one model of Ldpred2 for combining the results. I still cant figure out how to do that in correct way. Thank you so much for your prompt responses and help.

Best Regards, Mehreen

Get Outlook for Androidhttps://aka.ms/ghei36

From: Florian Privé notifications@github.com Sent: Tuesday, July 7, 2020 2:33:26 PM To: privefl/bigsnpr bigsnpr@noreply.github.com Cc: MehreenSoomro mahatyab@hotmail.com; Author author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

It should be fine to just sum the predictions from all 22 chromosomes.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://eur04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-654862829&data=02%7C01%7C%7Cc0106400b543450a0efa08d8227a545d%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637297256072130276&sdata=%2FtwneT7k%2BObcqZJ85ZZQ57O9ritajNcloaJLlI1h8T0%3D&reserved=0, or unsubscribehttps://eur04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYVTXWC2J5WILM6NW2TR2MP2NANCNFSM4NPENQJQ&data=02%7C01%7C%7Cc0106400b543450a0efa08d8227a545d%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637297256072130276&sdata=bDo9Q6sthiIlM%2B8R1KxuUADPX5AEdeggnN%2FysbDleSo%3D&reserved=0.

privefl commented 4 years ago

If you have PRS_chr1 for all individuals, PRS_chr2 for all individuals, and so on, you can compute PRS_chr1 + ... + PRS_chr22.

MehreenSoomro commented 4 years ago

Hi Florian,

Thanks for your reply.

I just wanted to ask, after summing up the result from all chromosome we get one PRS. So can you check for performance through AUC from this and if no how would you interpret it?

Best Regards, Mehreen Get Outlook for Androidhttps://aka.ms/ghei36

From: Florian Privé notifications@github.com Sent: Tuesday, July 7, 2020 2:42:58 PM To: privefl/bigsnpr bigsnpr@noreply.github.com Cc: MehreenSoomro mahatyab@hotmail.com; Author author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

If you have PRS_chr1 for all individuals, PRS_chr2 for all individuals, and so on, you can compute PRS_chr1 + ... + PRS_chr22.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-654870008&data=02%7C01%7C%7C7c7b283b6806419ebc6f08d8227ba973%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637297261797715522&sdata=HVyixOBJHp2qC2uBeUFvVPE3ZHNN4j9kA%2FZR4HrSsv8%3D&reserved=0, or unsubscribehttps://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYS57BNK2BYYBOSES53R2MQ6FANCNFSM4NPENQJQ&data=02%7C01%7C%7C7c7b283b6806419ebc6f08d8227ba973%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637297261797725516&sdata=eO5%2BOFJnToiZ%2BdS3Zofwbxhz0viAsh4MXJFokdFkUGo%3D&reserved=0.

privefl commented 4 years ago

Yes, this is the final PRS.

MehreenSoomro commented 4 years ago

How would you interpret that one PRS?

Get Outlook for Androidhttps://aka.ms/ghei36

From: Florian Privé notifications@github.com Sent: Saturday, July 11, 2020 4:10:42 PM To: privefl/bigsnpr bigsnpr@noreply.github.com Cc: MehreenSoomro mahatyab@hotmail.com; Author author@noreply.github.com Subject: Re: [privefl/bigsnpr] Computational time and memory (#87)

Yes, this is the final PRS.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fprivefl%2Fbigsnpr%2Fissues%2F87%23issuecomment-657078319&data=02%7C01%7C%7C2b2a5e5968834c981ad808d825ac94f0%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637300770440012180&sdata=FUHve%2BrrnjU8D2gt7E%2F6oVoMUu%2BF00ozLf9gQ7YVdJg%3D&reserved=0, or unsubscribehttps://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANQFWYT3TTWRKKL7KIW2VX3R3B6HFANCNFSM4NPENQJQ&data=02%7C01%7C%7C2b2a5e5968834c981ad808d825ac94f0%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637300770440022166&sdata=RjUhuaquO70aynM4wyrXHhz1r2xwHad%2BqAju7ReK5P8%3D&reserved=0.

privefl commented 4 years ago

I don't understand what you mean by "interpret". It is just a PRS that you use for anything you want, e.g. prediction.