FunVIP 0.3.20__py3-none-any.whl

src/trim.py

@@ -0,0 +1,118 @@
+from funvip.src import ext
+from funvip.src.opt_generator import opt_generator
+from Bio import AlignIO
+from Bio import SeqIO
+from time import sleep
+import multiprocessing as mp
+import os
+import logging
+import shutil
+
+
+## Trimming function of FunVIP
+# As most of the trimming functions affects inside of alignment
+def trimming(alignment, out, path, opt):
+    logging.debug("Entered trimming module")
+    # Save alignment before running trimming
+    original_msa = AlignIO.read(alignment, "fasta")
+
+    # Running trimming
+    if opt.method.trim.lower() == "gblocks":
+        trimming_result = ext.Gblocks(fasta=alignment, out=out, path=path)
+    elif opt.method.trim.lower() == "trimal":
+        trimming_result = ext.Trimal(
+            fasta=alignment,
+            out=out,
+            path=path,
+            algorithm=opt.trimal.algorithm,
+            threshold=opt.trimal.gt,
+        )
+    else:
+        trimming_result = shutil.copy(alignment, out)
+
+    logging.debug(f"opt.allow_innertrimming: {opt.allow_innertrimming}")
+    logging.debug(f"opt.method.trim: {opt.method.trim}")
+    # Repair mid alignment
+    if not (opt.allow_innertrimming) and (
+        opt.method.trim.lower()
+        in (
+            "gblocks",
+            "trimal",
+        )
+    ):
+        # Check if trimming successed
+        if trimming_result is not ((-1, -1)):
+            # Repair trimmend alignment by analysis flanking region
+            trimmed_msa = AlignIO.read(out, "fasta")
+            logging.debug(f"Trimming region for {alignment} : {trimming_result}")
+
+            # trimming results are in 1 based positions, and start, end included
+            revived_msa = original_msa[:, trimming_result[0] - 1 : trimming_result[1]]
+            AlignIO.write(revived_msa, out, "fasta")
+        else:
+            logging.warning(
+                f"Trimming {alignment} failed. Check if the alignment includes invalid sequneces"
+            )
+
+    return trimming_result
+
+
+# Trimming pipeline of FunVIP
+def pipe_trimming(V, path, opt):
+    trimming_opt = opt_generator(V, opt, path, step="trimming")
+    # run multiprocessing start
+    if opt.verbose < 3:
+        p = mp.Pool(opt.thread)
+        trimming_result = p.starmap(trimming, trimming_opt)
+        p.close()
+        p.join()
+
+    else:
+        # non-multithreading mode for debugging
+        trimming_result = []
+        for option in trimming_opt:
+            trimming_result.append(trimming(*option))
+
+    # sleep(10)
+    # raise Exception
+
+    # Remove datasets if trimming results nothing
+    trim_fail = []
+    for group in V.dict_dataset:
+        for gene in V.dict_dataset[group]:
+            if not gene == "concatenated":
+                if os.path.isfile(
+                    f"{path.out_alignment}/{opt.runname}_MAFFT_{group}_{gene}.fasta"
+                ):
+                    if os.path.isfile(
+                        f"{path.out_alignment}/{opt.runname}_trimmed_{group}_{gene}.fasta"
+                    ):
+                        seq_list = list(
+                            SeqIO.parse(
+                                f"{path.out_alignment}/{opt.runname}_trimmed_{group}_{gene}.fasta",
+                                "fasta",
+                            )
+                        )
+                        if len(seq_list[0].seq) == 0:
+                            trim_fail.append((group, gene))
+                    else:
+                        pass
+                else:
+                    pass
+
+    for fail in trim_fail:
+        group = fail[0]
+        gene = fail[1]
+        V.dict_dataset[group].pop(gene)
+        logging.warning(
+            f"Dataset {group} {gene} has removed during trimming because no sequence left. Please check alignment if database contains bad sequences."
+        )
+
+        # If non of the genes left for group except for concatenate, remove group
+        if len(V.dict_dataset[group]) == 1 and "concatenated" in V.dict_dataset[group]:
+            logging.warning(
+                f"Dataset {group} has removed because non of the genes are available"
+            )
+            V.dict_dataset.pop(group)
+
+    return V, path, opt