Merge pull request #1308 from SamuelSun2000/main

Remove redundant prodedures and columns

Author: gaow

code/reference_data/rss_ld_sketch.ipynb

@@ -768,7 +768,7 @@
     "    # Step 2: Merge all per-block pgens -> one per-chrom pgen\n",
     "    $[plink2_bin] \\\n",
     "        --pmerge-list \"${merge_list}\" pfile \\\n",
-    "        --make-pgen vzs \\\n",
+    "        --make-pgen \\\n",
     "        --sort-vars \\\n",
     "        --out  \"${final_prefix}\"\n",
     "\n",
@@ -787,16 +787,20 @@
     "    # Extract ID, MIN, and MAX from the concatenated .afreq (columns 2, 7, 8)\n",
     "    awk 'NR > 1 {print $2, $7, $8}' \"${final_prefix}.afreq\" > \"${final_prefix}.recovery_params\"\n",
     "\n",
-    "    # Decompress the .pvar.zst \n",
-    "    $[plink2_bin] --zst-decompress \"${final_prefix}.pvar.zst\" > \"${final_prefix}.pvar.tmp\"\n",
+    "    cp \"${final_prefix}.pvar\" \"${final_prefix}.pvar.tmp\"\n",
     "\n",
     "    awk 'NR==FNR {min[$1]=$2; max[$1]=$3; next} \n",
     "         /^#/ {print $0 \"\\tU_MIN\\tU_MAX\"; next} \n",
     "         {if ($3 in min) print $0 \"\\t\" min[$3] \"\\t\" max[$3]; else print $0 \"\\tNA\\tNA\"}' \\\n",
     "         \"${final_prefix}.recovery_params\" \"${final_prefix}.pvar.tmp\" > \"${final_prefix}.pvar\"\n",
     "\n",
     "    # Replace the compressed file with the updated text file\n",
-    "    rm \"${final_prefix}.pvar.tmp\" \"${final_prefix}.pvar.zst\" \"${final_prefix}.recovery_params\"\n",
+    "    rm \"${final_prefix}.pvar.tmp\" \"${final_prefix}.recovery_params\"\n",
+    "\n",
+    "    awk '{print $1,$2,$3,$4,$5,$6}' OFS=\"\\t\" \\\n",
+    "        \"${final_prefix}.afreq\" > \"${final_prefix}.afreq.tmp\"\n",
+    "    \n",
+    "    mv \"${final_prefix}.afreq.tmp\" \"${final_prefix}.afreq\"\n",
     "\n",
     "\n",
     "R: expand = \"$[ ]\"\n",