search

GMELab / MonsterLM

A method to estimate the proportion of variance explained by heritability and gene-by-environment interactions, in a fast, accurate, efficient, and unbiased manner on biobank-scale datasets (N>300,000).
GNU General Public License v3.0
1 stars 0 forks source link

Variable query #1

Open zhanji123 opened 1 year ago

zhanji123 commented 1 year ago

Hello, your research is very interesting, and I am very interested in your work. What I would like to inquire about is whether MonsterLM can incorporate multiple environmental exposures together when calculating, or if it can only include one environmental variable at a time?

discipim commented 7 months ago

Hi,

Apologies for the late reply. MonsterLM was designed and tested using environmental exposures of dimensionality n x 1. If you are suggesting using exposures of size n x 2 or greater then I think some further adaptation and simulation testing would be required.

However, MonsterLM can support environmental exposures of dimensionality n x 1 where the "E" is derived from multiple environments. For example, some fields will use the first principal component (PC1) as an "E".

zhanji123 commented 6 months ago

Thank you very much for your response, which is very helpful to us. However, the issue we're encountering with the update is that the program seems to be consuming excessive memory, leading to very slow overall command execution. We're dealing with over 300,000 individuals and 6,000,000 SNPs, and we're partitioning by chromosome. However, memory usage far exceeds 250GB, and the overall process is extremely slow. I'm not sure if this is normal, and I hope to get your answer.