KMDS(1)

💖 KMDS — control for population structure 💖

okay bestie, listen up, this is kmds, a little icon from the User Commands corner of the manual. she controls for population structure, filters kmers, and builds you a matrix that represents population structure. she's part of the seer(1) family, aka Sequence Element (kmer) Enrichment Analysis — so if you already know seer, you already kinda know her vibe. iykyk ✨

---

✨ DESCRIPTION ✨

so what does she actually do? Control for population structure. she filters kmers and creates a matrix representing population structure. that's the whole gig, and she eats every time 💅

she belongs to seer(1), the Sequence Element (kmer) Enrichment Analysis situation — so think of her as one of the main characters in that whole ensemble cast. 👑

---

🎀 OPTIONS 🎀

the two iconic workflows, no cap:

1. filter and subsample with --no_mds and --size
2. combine, and do metric multidimensional scaling with --mds_concat

---

🦋 Required options 🦋

these are non-negotiable, bestie. like, she literally will not go anywhere without them.

• -k, --kmers arg — the dsm kmer output file. this is your input, your whole reason for being here 🫶
• -p, --pheno arg — the .pheno metadata. the pheno file, the personality file, the lore.

---

🌸 MDS options 🌸

this is the multidimensional scaling section, and she is obsessed with it.

• -o, --output arg — output prefix for the new dsm file.
• --no_mds — do not perform MDS; instead she outputs the subsampled matrix. just the matrix, no scaling, no drama. 🙈
• --write_distances — writes a csv of the distance matrix. give the people the csv!! 📄
• --mds_concat arg — a list of subsampled matrices to use in MDS. note that this performs only MDS; it implies --no_filtering. so like, filtering is off the table the second you use this, periodt.
• --pc arg (=3) — number of principal coordinates to output. default is 3, which is honestly a serve.
• --size arg (=1000000) — number of kmers to use in MDS. default is 1000000 (a whole million, she's not doing anything small 💀).
• --threads arg (=1) — number of threads. default is 1. suggested: 4. so like, crank it up if your machine can handle the main character energy. 🔥

---

💜 Filtering options 💜

the filter era. very demure, very mindful.

• --no_filtering — turn off all filtering and do not output a new kmer file. everything stays, nothing new gets written. it's giving chaos but make it intentional.
• --max_length arg (=100) — maximum kmer length. default 100.
• --maf arg (=0.01) — minimum kmer frequency. default 0.01. (maf = minor allele frequency energy, keep it in mind 🧚)
• --min_words arg — minimum kmer occurrences. Overrides --maf. so if you set both, this one wins. she said "i'm the captain now". 👑

---

⭐ Other options ⭐

• --version — prints version and exits. one and done, bye 💅
• -h, --help — full help message. when in doubt, ask her.

---

🌈 EXAMPLE 🌈

want to filter kmers and create a matrix representing population structure with kmds? here's the move:

kmds -k dsm_input.txt.gz --pheno metadata.pheno -o filtered

and if you want to spread this process out, run the following command on each dsm file:

kmds -k dsm_input.txt.gz --pheno metadata.pheno --no_mds --size 10000

copy it exactly, bestie, don't freestyle the flags. 😭

---

🍓 AUTHOR 🍓

this manpage was written by Andreas Tille for the Debian distribution, and can be used for any other usage of the program. shoutout to Andreas fr fr 🙌

---

ok that's the whole page!! everything she does, every flag, every default, all of it. no cap. 💖

and a lil sign-off for the girls and the gays and everyone in between: this is kmds version 1.1.2, dated March 2017. vintage but she still slays. 🎀✨🦋