SMARTEIGENSTRAT(1)

✨ NAME ✨

smarteigenstrat — reduction of population bias for genetic analyses 💅

ok bestie, real talk: this is the manual page for smarteigenstrat, living in section 1 (User Commands) of your friendly neighbourhood man system. we're gonna read it together, i'll hold your hand, it's gonna be iconic 🫶

🦋 SYNOPSIS 🦋

smarteigenstrat -i <genotypefile>  -a <indivfile>  -b <snpfile>  -p
<pcafile>   -o <outputfile>      -l <logfile>  -k <numpca>  -q <qtlmode>

like yeah, that's a lot of flags and zero breathing room, but don't panic — we go through every single one below, no cap 🫡

💖 DESCRIPTION 💖

so the program smarteigenstrat is part of EIGENSOFT. that's the whole family, she's one of the siblings.

the EIGENSOFT package combines functionality from the group's population genetics methods (Patterson et al. 2006) and their EIGENSTRAT stratification method (Price et al. 2006). 🧬

and the EIGENSTRAT method? she uses principal components analysis to explicitly model ancestry differences between cases and controls along continuous axes of variation — the way that it does this is so specific, honestly. the resulting correction is specific to a candidate marker's variation in frequency across ancestral populations, which minimizes spurious associations while maximizing power to detect true associations. that's the whole glow-up: less fake noise, more real signal. periodt. 💅

also the EIGENSOFT package has a built-in plotting script (slay) and supports multiple file formats and quantitative phenotypes 🌈

🎀 OPTIONS 🎀

every flag, every single one, we leave nobody behind:

• -i — genotype file (PED, PACKEDPED, EIGENSTRAT, ANCESTRYMAP or PACKEDANCESTRYMAP format) 🧬
• -a — the indivfile, that's it, that's the description the manpage gives us, we respect her brevity 🫶
• -b — the snpfile, same energy, same honesty ✨
• -p — the pcafile 💜
• -o — output file (chisq) 🔥
• -l — logfile (screen output, including error messages) — yes, including error messages, so when something goes wrong, this is where you'll find out 😭
• -k — the numpca value 🎀
• -q — YES for quantitative phenotype or NO otherwise

and now the ✨phenotype lore✨ because it matters, bestie:

For quantitative phenotype, the sixth column of .ped file or the third column of EIGENSTRAT .ind file should be real numbers. 🔢

For non-quantitative phenotype, that same sixth column of .ped or third column should be 'Case' or 'Control'. 👑

ngl if you put the wrong kind of value in those columns, that's on you — the manpage is very clear about which is which. double-check before you run it fr fr 🙈

👑 AUTHOR 👑

This manpage was written by Andreas Tille for the Debian distribution and can be used for any other usage of the program. thank you Andreas, king behaviour, we appreciate you 💖

and that's the whole page, bestie 🦋 — this was smarteigenstrat 6.1.2, dated July 2016, and it ate, no crumbs. go run your PCA, you've got this 🌸✨