fix ENC

cubat2/.ipynb_checkpoints/Untitled-checkpoint.ipynb

@@ -0,0 +1,37 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "45bf49b1-e761-4502-b5b9-32befcab53ca",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "library(ggplot2)\n",
+    "library(data.table)\n",
+    "library(Bioconductor::Biostrings)\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

cubat2/.ipynb_checkpoints/test-checkpoint.r

@@ -0,0 +1,3 @@
+library(ggplot2)
+library(data.table)
+library(Bioconductor::Biostrings)

cubat2/Untitled.ipynb

@@ -0,0 +1,54 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "45bf49b1-e761-4502-b5b9-32befcab53ca",
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "SyntaxError",
+     "evalue": "invalid syntax (2852074393.py, line 3)",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[0;36m  Cell \u001b[0;32mIn[1], line 3\u001b[0;36m\u001b[0m\n\u001b[0;31m    library(Bioconductor::Biostrings)\u001b[0m\n\u001b[0m                        ^\u001b[0m\n\u001b[0;31mSyntaxError\u001b[0m\u001b[0;31m:\u001b[0m invalid syntax\n"
+     ]
+    }
+   ],
+   "source": [
+    "library(ggplot2)\n",
+    "library(data.table)\n",
+    "library(Bioconductor::Biostrings)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "52eee6d8-e483-47a2-9efe-b033d22c4f2d",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

cubat2/codon_table.py

@@ -1,3 +1,4 @@
+import pandas as pd
 from Bio.Data import CodonTable
 
 
@@ -57,8 +58,37 @@ def codon_table_completion(index=1):
     return complete_codon_table
 
 
-# 用法示例：
+def codon_table_subfamily(index= 1):
+    """
+    根据密码子表编号获取密码子表的子家族信息。
+
+    参数:
+    - index (str): 密码子表的编号，默认为 '1'。
+
+    返回:
+    - codon_table_subfamily (pd.DataFrame): 包含密码子子家族信息的DataFrame, 包括 aa_code, amino_acid, codon, 和 subfam 列。
+    """
+    codon_table = CodonTable.unambiguous_dna_by_id[int(index)]
+    codon_table_dict = {
+        'codon': [],
+        'amino_acid': [],
+        'subfamily': []
+    }
+
+    for codon, amino_acid in codon_table.forward_table.items():
+        aa_code = CodonTable.ambiguous_dna_by_id[int(index)].forward_table[codon]
+        subfamily = f"{amino_acid}_{codon[:2]}"
+        codon_table_dict['codon'].append(codon)
+        codon_table_dict['amino_acid'].append(amino_acid)
+        codon_table_dict['subfamily'].append(subfamily)
+
+    codon_table_subfamily = pd.DataFrame(codon_table_dict)
+    return codon_table_subfamily
+
+
+# 测试代码：
 if __name__ == "__main__":
     index = 1  # 例如，选择标准密码子表
-    codon_table = get_codon_table_by_index(index)
+    codon_table = codon_table_subfamily(index)
     print(codon_table)
+

cubat2/example.csv

@@ -0,0 +1,116 @@
+"type","codon","anticodon"
+"WC","TTG","CAA"
+"WC","TTA","TAA"
+"WC","TTC","GAA"
+"IU","TTT","AAA"
+"IC","TTC","AAA"
+"GU","TTT","GAA"
+"UG","TTG","TAA"
+"WC","TCG","CGA"
+"WC","TCA","TGA"
+"WC","TCC","GGA"
+"IU","TCT","AGA"
+"IC","TCC","AGA"
+"IA","TCA","AGA"
+"GU","TCT","GGA"
+"UG","TCG","TGA"
+"WC","TAC","GTA"
+"IU","TAT","ATA"
+"IC","TAC","ATA"
+"GU","TAT","GTA"
+"WC","TGG","CCA"
+"WC","TGC","GCA"
+"IU","TGT","ACA"
+"IC","TGC","ACA"
+"GU","TGT","GCA"
+"WC","CTG","CAG"
+"WC","CTA","TAG"
+"WC","CTC","GAG"
+"IU","CTT","AAG"
+"IC","CTC","AAG"
+"IA","CTA","AAG"
+"GU","CTT","GAG"
+"UG","CTG","TAG"
+"WC","CCG","CGG"
+"WC","CCA","TGG"
+"WC","CCC","GGG"
+"IU","CCT","AGG"
+"IC","CCC","AGG"
+"IA","CCA","AGG"
+"GU","CCT","GGG"
+"UG","CCG","TGG"
+"WC","CAG","CTG"
+"WC","CAA","TTG"
+"WC","CAC","GTG"
+"IU","CAT","ATG"
+"IC","CAC","ATG"
+"GU","CAT","GTG"
+"UG","CAG","TTG"
+"WC","CGG","CCG"
+"WC","CGA","TCG"
+"WC","CGC","GCG"
+"IU","CGT","ACG"
+"IC","CGC","ACG"
+"IA","CGA","ACG"
+"GU","CGT","GCG"
+"UG","CGG","TCG"
+"WC","ATG","CAT"
+"WC","ATA","TAT"
+"WC","ATC","GAT"
+"IU","ATT","AAT"
+"IC","ATC","AAT"
+"IA","ATA","AAT"
+"GU","ATT","GAT"
+"WC","ACG","CGT"
+"WC","ACA","TGT"
+"WC","ACC","GGT"
+"IU","ACT","AGT"
+"IC","ACC","AGT"
+"IA","ACA","AGT"
+"GU","ACT","GGT"
+"UG","ACG","TGT"
+"WC","AAG","CTT"
+"WC","AAA","TTT"
+"WC","AAC","GTT"
+"IU","AAT","ATT"
+"IC","AAC","ATT"
+"GU","AAT","GTT"
+"UG","AAG","TTT"
+"WC","AGG","CCT"
+"WC","AGA","TCT"
+"WC","AGC","GCT"
+"IU","AGT","ACT"
+"IC","AGC","ACT"
+"GU","AGT","GCT"
+"UG","AGG","TCT"
+"WC","GTG","CAC"
+"WC","GTA","TAC"
+"WC","GTC","GAC"
+"IU","GTT","AAC"
+"IC","GTC","AAC"
+"IA","GTA","AAC"
+"GU","GTT","GAC"
+"UG","GTG","TAC"
+"WC","GCG","CGC"
+"WC","GCA","TGC"
+"WC","GCC","GGC"
+"IU","GCT","AGC"
+"IC","GCC","AGC"
+"IA","GCA","AGC"
+"GU","GCT","GGC"
+"UG","GCG","TGC"
+"WC","GAG","CTC"
+"WC","GAA","TTC"
+"WC","GAC","GTC"
+"IU","GAT","ATC"
+"IC","GAC","ATC"
+"GU","GAT","GTC"
+"UG","GAG","TTC"
+"WC","GGG","CCC"
+"WC","GGA","TCC"
+"WC","GGC","GCC"
+"IU","GGT","ACC"
+"IC","GGC","ACC"
+"IA","GGA","ACC"
+"GU","GGT","GCC"
+"UG","GGG","TCC"