📢 Interested in new phenotypes for feed efficiency in aquaculture? 📢 We are hiring a fully salaried PhD in genetics of FE in Atlantic salmon 🧬🐟 @UniNMBU 🇳🇴 #PhD#PhenotypeIsKing. Please RT. 📅 deadline 5 June 24 jobbnorge.no/en/available-jo…
Are you our new collegue? We are hiring a fully fund PhD in simulation tools for managing genetic diversity 🧬🐄🐟 @UniNMBU 🇳🇴 #PhD.Please RT! jobbnorge.no/en/available-jo…
It is very nice to have @GregorGorjanc to talk about his AlphaSimR in Ås. In the end he praised tree sequence storage a lot. This reminded me that I had a similar idea 11 yrs ago (xijiang.blogspot.com/2011/11…). Unfortunately, I still don't know graph theory.
To count similarities between two loci, one can code genotype 0, 1, and 2 as bits 00, 01, and 11. Two CPU cycles, xor popcnt, can count 32 loci. A 2Ghz CPU can count 6.4x10^10 loci. Or, 32k ID, 50k loci in 20m.
Managed to deal with SNP matrix(50k x 400k), LHS of 50k x 50k with only ~20G peak memory usage. Paralleled of course. W/O the @JuliaLanguage, it can't be this compact and this fast.
English speakers usually insert 'you know' blah, blah when talking. The similar Chinese words are "这个/那个", meaning this or that. The latter sounds like the n-word. So if some Chinese stutters n-word, it doesn't mean that. It only means that 'that'. It is not racist at all.