identical result from GLM and FarmCPU

[binlu1981]

Hi @XiaoleiLiuBio @YinLiLin @hyacz ,
I got exactly the same p-values from GLM and FarmCPU, but different results from MLM method. The rMVP version is 1.06 and set method.bin="static". Is there a bug or some problems?

Best
Bin

[XiaoleiLiuBio]

Hi Bin, Thank you for using rMVP. If there is no signal passed the Bonferroni cutoff in GLM, FarmCPU will stop and output the same results as GLM. Best, Xiaolei

----------------

---------- Origin message ----------

From："Bin Lu" ***@***.***> To："xiaolei-lab/rMVP" ***@***.***> Subject：[xiaolei-lab/rMVP] identical result from GLM and FarmCPU (Issue #72) Date：2022-02-08 14:41:32Hi @XiaoleiLiuBio @YinLiLin @hyacz ,

I got exactly the same p-values from GLM and FarmCPU, but different results from MLM method. The rMVP version is 1.06 and set method.bin="static". Is there a bug or some problems? Best Bin

…

-- Reply to this email directly or view it on GitHub: #72 You are receiving this because you were mentioned. Message ID: ***@***.***>

[binlu1981]

Hi @XiaoleiLiuBio ,
Thanks for quick reply. I have nearly 30 million snps, so the correction cutoff seems too stringent for the data. How can I make FarmCPU work? Should I set p.threshold higher or change some other parameter?
one more question. Which method.bin is prefered? FaST-LMM or static? Look forward to see your suggestions.

Best
Bin

[XiaoleiLiuBio]

Yes, you can change the p.threshold to an appropriate cutoff to make FarmCPU work but should be very careful. The default of p.threshold is set to Bonferroni cutoff.

----------------

---------- Origin message ----------

From："Bin Lu" ***@***.***> To："xiaolei-lab/rMVP" ***@***.***> Subject：Re: [xiaolei-lab/rMVP] identical result from GLM and FarmCPU (Issue #72) Date：2022-02-08 22:30:11Hi @XiaoleiLiuBio ,

Thanks for quick reply. I have nearly 30 million snps, so the correction cutoff seems too stringent for the data. How can I make FarmCPU work? Should I set p.threshold higher or change some other parameter? Best Bin

…

-- Reply to this email directly or view it on GitHub: #72 (comment) You are receiving this because you were mentioned. Message ID: ***@***.***>

[binlu1981]

Thanks Xiaolei. One more question. Which method for selecting the most appropriate bins is preffered, ’static’, ’EMMA’ or ’FaST-LMM’?

Best
Bin

[XiaoleiLiuBio]

The default is set to 'static' for the fastest speed and 'FaST-LMM' is recommended for stable Power/FDR performance.

----------------

---------- Origin message ----------

From："Bin Lu" ***@***.***> To："xiaolei-lab/rMVP" ***@***.***> Subject：Re: [xiaolei-lab/rMVP] identical result from GLM and FarmCPU (Issue #72) Date：2022-02-11 15:42:34Thanks Xiaolei. One more question. Which method for selecting the most appropriate bins is preffered, ’static’, ’EMMA’ or ’FaST-LMM’?

Best Bin

…

-- Reply to this email directly or view it on GitHub: #72 (comment) You are receiving this because you were mentioned. Message ID: ***@***.***>