如何在DataFrame中拆分基因变异字符串，提取位置与变异类型

Hey Justin, let's figure out how to split those variant strings into the two columns you need. First, let's break down the structure of your entries:

• Most start with a reference ID like NC_000001.10:g., followed by a numeric position, then a mutation like G>C
• One entry uses m. instead of g., so we'll make sure our solution handles both cases.

The Right Regular Expression #

Here's a regex that will reliably capture both the position and mutation type for all your entries:

.*\.(g|m)\.(\d+)([A-Z]>[A-Z])

Let's break down what each part does:

• .*\.(g|m)\.: Matches everything up to (and including) the g. or m. prefix. The \. escapes the literal dot (since dots in regex match any character by default).
• (\d+): Captures one or more digits — this is your position number.
• ([A-Z]>[A-Z]): Captures the mutation pattern (e.g., G>C, A>G) by matching an uppercase letter, followed by >, followed by another uppercase letter.

Example Implementation (Python) #

If you're working with Python, here's how you can use this regex to process your dataset:

import re

# Your sample data
variant_list = [
    "NC_000001.10:g.955563G>C",
    "NC_000001.10:g.955597G>T",
    "NC_000001.10:g.955619G>C",
    "NC_000001.10:g.957640C>T",
    "NC_000001.10:g.976059C>T",
    "NC_000003.11:g.37090470C>T",
    "NC_000012.11:g.133256600G>A",
    "NC_012920.1:m.15923A>G"
]

# Our regex pattern
pattern = r".*\.(g|m)\.(\d+)([A-Z]>[A-Z])"

for variant in variant_list:
    match = re.fullmatch(pattern, variant)
    if match:
        position = match.group(2)  # Group 2 holds the numeric position
        mutation = match.group(3)  # Group 3 holds the mutation type
        print(f"Position: {position}, Mutation: {mutation}")

Running this will output:

Position: 955563, Mutation: G>C
Position: 955597, Mutation: G>T
Position: 955619, Mutation: G>C
Position: 957640, Mutation: C>T
Position: 976059, Mutation: C>T
Position: 37090470, Mutation: C>T
Position: 133256600, Mutation: G>A
Position: 15923, Mutation: A>G

For Spreadsheet Tools (Like Excel/Google Sheets) #

If you're using a spreadsheet instead of code, you can use these formulas:

• Extract Position: =MID(A1, FIND(".", A1, FIND(".", A1)+1)+1, FIND(">", A1)-FIND(".", A1, FIND(".", A1)+1)-1)
• Extract Mutation Type: =MID(A1, FIND(">", A1)-1, 3)

Just replace A1 with the cell containing your variant string.

内容的提问来源于stack exchange，提问作者Justin