This lesson is only available in Korean for now.
Dr. Seonggu Sohn's Python, First Steps

Make the paper's numbers
with your own hands

The 11,070 to 3,055 in the paper, and the 961 genes โ€” someone made those with code.
These 20 lessons cover just enough Python to read and write that code, starting from variables and ending by producing those very numbers.

๐Ÿ No install๐Ÿ“Š pandas๐ŸŒ‹ Volcano ๐Ÿงช Reproduce DDX54๐Ÿ‡ฌ๐Ÿ‡ง English
Start lesson 1 โ†’

How you'll learn

๐Ÿ“ฆNothing to install

Python runs inside your browser. Press โ–ถ and it executes right there โ€” edit the code and press it again.

๐ŸงฌEvery example is biology

Sequences, expression values and p-values instead of shopping carts and grade books. What you learn is what you will use.

๐ŸŽ›๏ธMove the cut-offs yourself

Grab the sliders and shake 0.3 and 0.05. You watch the number of surviving genes change.

๐Ÿ“„Reproduce the paper at the end

In lesson 19 you re-apply the paper's four selection steps in code, narrowing 11,070 down to 3,055.

Full curriculum

01
์™œ ํŒŒ์ด์ฌ์ธ๊ฐ€ โ€” ๋…ผ๋ฌธ์˜ ์ˆซ์ž๋ฅผ ๋‚ด ์†์œผ๋กœ
์ด 20๊ฐ•์ด ๋๋‚˜๋ฉด ๋…ผ๋ฌธ SI ์˜ 11,070 โ†’ 3,055 ๋ฅผ ์ง์ ‘ ๋งŒ๋“ ๋‹ค
โœ“
02
๊ฐ’๊ณผ ์ด๋ฆ„ํ‘œ โ€” ๋ณ€์ˆ˜ยท์ˆซ์žยท์ฐธ๊ฑฐ์ง“
TMB = ๋Œ์—ฐ๋ณ€์ด ์ˆ˜ รท 38 ์„ ์ ์„ ์ˆ˜ ์žˆ๊ฒŒ
โœ“
03
๋ฌธ์ž์—ด โ€” ์„œ์—ด์€ ๊ธ€์ž์˜ ์ค„์ด๋‹ค
๋…ผ๋ฌธ Table S1ยทS3 ์˜ siRNAยทguide ์„œ์—ด์„ ๋‹ค๋ค„ ๋ณธ๋‹ค
โœ“
04
๋ฆฌ์ŠคํŠธ โ€” ์œ ์ „์ž ๋ชฉ๋ก์„ ๋‹ด๋Š”๋‹ค
๋…ผ๋ฌธ SI ใ€Œ40 genes related to the antigen-presenting pathwayใ€
โœ“
05
๋”•์…”๋„ˆ๋ฆฌ โ€” ์œ ์ „์ž์—์„œ ๊ฐ’์„ ์ฐพ๋Š”๋‹ค
์ด๋ฆ„์œผ๋กœ ์ฐพ๊ธฐ ยท ๋ชฉ๋ก์—์„œ ํ‘œ๋กœ
โœ“
06
์กฐ๊ฑด โ€” p < 0.05 ๋ฅผ ์ฝ”๋“œ๋กœ ์ ๋Š”๋‹ค
๋…ผ๋ฌธ SI ใ€Œabsolute log2 fold change > 0.3 with P-value < 0.05ใ€
โœ“
07
ํ•จ์ˆ˜ โ€” ๋˜ํ’€์ด๋ฅผ ์ด๋ฆ„์œผ๋กœ ๋ฌถ๋Š”๋‹ค
๊ฐ™์€ ๊ณ„์‚ฐ์„ ์„ธ ๋ฒˆ ์“ฐ๊ธฐ ์ „์—
โœ“
08
FASTA ๋ฅผ ์ฝ๋Š”๋‹ค โ€” ๊ฐ€์žฅ ์˜ค๋ž˜๋œ ์ƒ๋ฌผํ•™ ํŒŒ์ผ
> ๋กœ ์‹œ์ž‘ํ•˜๋Š” ์ค„์ด ์ด๋ฆ„, ๋‚˜๋จธ์ง€๊ฐ€ ์„œ์—ด
โœ“
09
ํ‘œ๋ฅผ ์ฝ๋Š”๋‹ค โ€” CSV ์™€ TSV
๋ฐœํ˜„๋Ÿ‰ ํ‘œ๋Š” ๊ฒฐ๊ตญ ์‰ผํ‘œ(๋˜๋Š” ํƒญ)๋กœ ๋‚˜๋‰œ ๊ธ€์ž์ž…๋‹ˆ๋‹ค
โœ“
10
numpy โ€” ์ˆซ์ž ๋ฉ์–ด๋ฆฌ๋ฅผ ํ•œ๊บผ๋ฒˆ์—
2๋งŒ ๊ฐœ ์œ ์ „์ž์— ๋ฐ˜๋ณต๋ฌธ์„ ์“ฐ์ง€ ์•Š๋Š” ๋ฒ•
โœ“
11
pandas โ€” ์ด๋ฆ„์ด ๋ถ™์€ ํ‘œ
R ์˜ data.frame ์— ํ•ด๋‹นํ•˜๋Š” ๊ฒƒ
โœ“
12
ํ‘œ๋ฅผ ์ž‡๋Š”๋‹ค โ€” merge ์™€ groupby
๋ฐœํ˜„๋Ÿ‰ ํ‘œ + ํ‘œ๋ณธ ์ •๋ณด ํ‘œ = ๋ถ„์„ํ•  ์ˆ˜ ์žˆ๋Š” ํ‘œ
โœ“
13
matplotlib โ€” ์ˆซ์ž๋ฅผ ๊ทธ๋ฆผ์œผ๋กœ
๋…ผ๋ฌธ ๊ทธ๋ฆผ์€ ๊ฒฐ๊ตญ ๋ช‡ ์ค„์˜ ์ฝ”๋“œ์ž…๋‹ˆ๋‹ค
โœ“
14
์™œ log2 ๋ฅผ ์”Œ์šฐ๋‚˜ โ€” ๋ฐœํ˜„๋Ÿ‰์˜ ๋ˆˆ๊ธˆ
๋…ผ๋ฌธ์˜ ใ€Œlog2 fold changeใ€ ๊ฐ€ ๋กœ๊ทธ์ธ ์ด์œ 
โœ“
15
scipy โ€” t-๊ฒ€์ •์„ ํ•œ ์ค„๋กœ
๋…ผ๋ฌธ SI ใ€ŒWilcoxon rank-sum test or t-test were usedใ€
โœ“
16
๋ณผ์ผ€์ด๋…ธ ํ”Œ๋กฏ โ€” ์ฐจ๋“ฑ๋ฐœํ˜„์„ ํ•œ ์žฅ์—
๊ฐ€๋กœ๋Š” log2FC, ์„ธ๋กœ๋Š” โˆ’log10(p)
โœ“
17
ํžˆํŠธ๋งต โ€” ๋…ผ๋ฌธ Fig. 3 ์˜ pheatmap ์„ ํŒŒ์ด์ฌ์œผ๋กœ
SI ใ€Œwe employed the pheatmap package in R to produce heatmapsใ€
โœ“
18
๋ง‰ํ˜”์„ ๋•Œ โ€” ์˜ค๋ฅ˜๋ฅผ ์ฝ๋Š” ๋ฒ•
์˜ค๋ฅ˜ ๋ฉ”์‹œ์ง€๋Š” ๊พธ์ค‘์ด ์•„๋‹ˆ๋ผ ์•ˆ๋‚ด์ž…๋‹ˆ๋‹ค
โœ“
19
ใ€์‹ค์Šตใ€‘ ๋…ผ๋ฌธ์˜ 11,070 โ†’ 3,055 ๋ฅผ ๋งŒ๋“ ๋‹ค
SI ใ€ŒSample selection from TCGA dataใ€ ยท Fig. S1
โœ“
20
์ •๋ฆฌ โ€” ์—ฌ๊ธฐ์„œ ์–ด๋””๋กœ
R ๋„๊ตฌ์— ๋Œ€์‘ํ•˜๋Š” ํŒŒ์ด์ฌ ยท ๋‹ค์Œ ๊ณผ๋ชฉ
โœ“