From 31d4e89a32e2eb9c540b2f35728f26d518c06c0c Mon Sep 17 00:00:00 2001
From: Pjotr Prins
Date: Wed, 28 Aug 2024 10:22:29 +0200
Subject: gemma-wrapper: comment on GRM reduce

---
 topics/lmms/gemma/permutations.gmi | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'topics/lmms/gemma/permutations.gmi')

diff --git a/topics/lmms/gemma/permutations.gmi b/topics/lmms/gemma/permutations.gmi
index 9559366..3e2326d 100644
--- a/topics/lmms/gemma/permutations.gmi
+++ b/topics/lmms/gemma/permutations.gmi
@@ -351,6 +351,8 @@ So, the idea is to rerun permutations with the small set, but with the reduced G
 
 The interesting bit is that GEMMA requires input of phenotypes, but does not use them to compute the GRM.
 
+After giving it some thought we want GRM reduction to work in production GN because of the speed benefit. That means modifying gemma-wrapper to take a list of genometypes as input - and we'll output that with GN. It is a good idea anyhow because it can give us some improved error feedback down the line.
+
 WIP
 
 ## Dealing with epoch
-- 
cgit v1.2.3