darkprofiler 0.2.2__tar.gz → 0.2.3__tar.gz

{darkprofiler-0.2.2/src/darkprofiler.egg-info → darkprofiler-0.2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: darkprofiler
-Version: 0.2.2
+Version: 0.2.3
 Summary: DarkProfiler: Alignment and Classification of Peptides from Reference-Independent De Novo Peptide Sequencing Experiments.
 Author-email: Hanjun Lee <hanjun@alum.mit.edu>
 License: MIT
@@ -244,7 +244,6 @@ The database contains translated and derived proteomes as FASTA files:
 - `mutanome.fa`
 - `mutatedCanonicalTranscriptome.fa`
 - `mutatedAlternativeTranslatome.fa`
-- `mutatedAlternativeORFeome.fa`
 
 DarkProfiler also creates **persistent fast indices** under the same database directory to accelerate peptide search with Hamming distance:
 for example:
@@ -361,7 +360,7 @@ classify_peptides(
 - **ORF region labels**  
   For alternative ORF hits, DarkProfiler labels the peptide start as:
   - `uORF` (upstream of CDS start)
-  - `intORF` (inside annotated CDS span)
+  - `intORF` (out-of-frame peptdies from inside annotated CDS span)
   - `dORF` (downstream of CDS end)
   - `lncRNA` (no CDS annotation)
 

{darkprofiler-0.2.2 → darkprofiler-0.2.3}/README.md

@@ -226,7 +226,6 @@ The database contains translated and derived proteomes as FASTA files:
 - `mutanome.fa`
 - `mutatedCanonicalTranscriptome.fa`
 - `mutatedAlternativeTranslatome.fa`
-- `mutatedAlternativeORFeome.fa`
 
 DarkProfiler also creates **persistent fast indices** under the same database directory to accelerate peptide search with Hamming distance:
 for example:
@@ -343,7 +342,7 @@ classify_peptides(
 - **ORF region labels**  
   For alternative ORF hits, DarkProfiler labels the peptide start as:
   - `uORF` (upstream of CDS start)
-  - `intORF` (inside annotated CDS span)
+  - `intORF` (out-of-frame peptdies from inside annotated CDS span)
   - `dORF` (downstream of CDS end)
   - `lncRNA` (no CDS annotation)
 

{darkprofiler-0.2.2 → darkprofiler-0.2.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "darkprofiler"
-version = "0.2.2"
+version = "0.2.3"
 description = "DarkProfiler: Alignment and Classification of Peptides from Reference-Independent De Novo Peptide Sequencing Experiments."
 readme = "README.md"
 requires-python = ">=3.7"

{darkprofiler-0.2.2 → darkprofiler-0.2.3}/src/darkprofiler/__init__.py

@@ -2,5 +2,5 @@ from .run import classify_peptides
 
 __all__ = ["classify_peptides"]
 
-__version__ = "0.2.2"
+__version__ = "0.2.3"
 

{darkprofiler-0.2.2 → darkprofiler-0.2.3}/src/darkprofiler/cli.py

@@ -121,7 +121,7 @@ def build_parser() -> argparse.ArgumentParser:
             "Optional path to existing database directory containing "
             "canonicalProteome.fa, alternativeSplicing.fa, mutanome.fa, "
             "mutatedCanonicalTranscriptome.fa, mutatedAlternativeTranslatome.fa, "
-            "mutatedAlternativeORFeome.fa."
+            "and other index files"
         ),
     )
     p_run.add_argument(

{darkprofiler-0.2.2 → darkprofiler-0.2.3}/src/darkprofiler/run.py

@@ -481,7 +481,6 @@ def classify_peptides(reference,
 
     required_db_files = [
         "alternativeSplicing.fa",
-        "mutatedAlternativeORFeome.fa",
         "canonicalProteome.fa",
         "mutatedAlternativeTranslatome.fa",
         "mutanome.fa",
@@ -1006,31 +1005,169 @@ def classify_peptides(reference,
 
     report_step_done()  # 5
 
-    # ----------------------------- mutated antigens (mutanome) -------------------
-    # NOTE: keep your original SNV application as-is (omitted here for brevity),
-    # but ensure translation uses translate_cds_with_map (ATG filter) and then build index.
-    #
-    # If you already have your SNV block, set:
-    #   mutated_canonical_transcripts = load_transcriptome(mutatedCanonicalTranscriptome.fa)
-    #   mutanome_proteins, mutanome_aa2nt = translate_cds_with_map(mutated_canonical_transcripts, cds_map)
-    #   write mutanome.fa
-    #
+    # ---------- build mutanome.fa + mutatedCanonicalTranscriptome.fa ----------
     mutated_canonical_tx_fa = os.path.join(database_dir, "mutatedCanonicalTranscriptome.fa")
     mutanome_fa = os.path.join(database_dir, "mutanome.fa")
     mutanome_idx_dir = os.path.join(database_dir, "mutanome.idx")
 
-    # ---------- BEGIN: your existing SNV block should fill these ----------
-    # For runs reusing DB, mutated_canonical_tx_fa and mutanome_fa should already exist.
-    mutated_canonical_transcripts = {}
-    if os.path.exists(mutated_canonical_tx_fa):
-        mutated_canonical_transcripts = load_transcriptome(mutated_canonical_tx_fa)
+    def _parse_gff_attributes(attr_str):
+        attrs = {}
+        for item in attr_str.strip().split(";"):
+            item = item.strip()
+            if not item:
+                continue
+            if "=" in item:
+                k, v = item.split("=", 1)
+                attrs[k] = v.strip().strip('"')
+            else:
+                parts = item.split()
+                if len(parts) >= 2:
+                    attrs[parts[0]] = parts[1].strip('"')
+        return attrs
+
+    # Build transcript exon model from GFF (exon features only)
+    transcript_exons = defaultdict(list)    # tx_id -> list[(start,end)] 1-based closed
+    transcript_strand = {}                 # tx_id -> '+'/'-'
+    transcript_chrom = {}                  # tx_id -> chrom (normalized)
+
+    with open(gff_path) as fh:
+        for line in fh:
+            if not line.strip() or line.startswith("#"):
+                continue
+            fields = line.rstrip("\n").split("\t")
+            if len(fields) < 9:
+                continue
+            chrom, source, feature, start, end, score, strand, frame, attrs_str = fields
+            if feature.lower() != "exon":
+                continue
+            try:
+                start_i = int(start)
+                end_i = int(end)
+            except ValueError:
+                continue
+            attrs = _parse_gff_attributes(attrs_str)
+            tx_id = attrs.get("transcript_id") or attrs.get("transcriptId")
+            if tx_id is None:
+                continue
+            tx_id = normalize_gff_tx_id(tx_id)
+            transcript_exons[tx_id].append((start_i, end_i))
+            transcript_strand[tx_id] = strand
+            transcript_chrom[tx_id] = normalize_chrom(chrom)
+
+    # Sort exons ascending by genomic start
+    for tx_id in list(transcript_exons.keys()):
+        transcript_exons[tx_id].sort(key=lambda x: x[0])
+
+    # Exon caches for coordinate mapping
+    exon_order_cache = {}  # tx_id -> (exons_sorted, exons_desc)
+    for tx_id, exons_sorted in transcript_exons.items():
+        exon_order_cache[tx_id] = (exons_sorted, list(reversed(exons_sorted)))
+
+    # Index SNVs by chromosome for speed
+    snvs_by_chrom = defaultdict(list)  # chrom -> list[(pos, ref, alt)]
+    for chrom, pos, ref, alt in snvs:
+        snvs_by_chrom[chrom].append((int(pos), ref, alt))
+    for chrom in snvs_by_chrom:
+        snvs_by_chrom[chrom].sort(key=lambda x: x[0])
+
+    # Simple base complement
+    _complement = {"A":"T","T":"A","C":"G","G":"C","a":"t","t":"a","c":"g","g":"c"}
+    def _complement_base(b):
+        return _complement.get(b, b)
+
+    def _apply_snvs_to_one_transcript(tx_id):
+        """
+        Returns: (tx_id, mutated_seq_string) OR (tx_id, None) if transcript not present.
+        If no SNVs (or no mapping), returns original transcript sequence.
+        """
+        if tx_id not in transcriptome:
+            return tx_id, None
+
+        seq = transcriptome[tx_id]
+        seq_list = list(seq)
+
+        chrom = transcript_chrom.get(tx_id)
+        if chrom is None or chrom not in snvs_by_chrom:
+            return tx_id, seq
+
+        if tx_id not in exon_order_cache:
+            return tx_id, seq
+
+        exons_sorted, exons_desc = exon_order_cache[tx_id]
+        strand = transcript_strand.get(tx_id, "+")
+
+        for pos, ref, alt in snvs_by_chrom[chrom]:
+            if strand == "+":
+                offset = 0
+                within = False
+                for s, e in exons_sorted:
+                    if pos < s:
+                        break
+                    if pos > e:
+                        offset += (e - s + 1)
+                    else:
+                        offset += (pos - s)
+                        within = True
+                        break
+                if not within:
+                    continue
+                tx_index = offset
+                expected_ref = ref.upper()
+                alt_base = alt.upper()
+            else:
+                offset = 0
+                within = False
+                for s, e in exons_desc:
+                    if pos > e:
+                        continue
+                    if pos < s:
+                        offset += (e - s + 1)
+                    else:
+                        offset += (e - pos)
+                        within = True
+                        break
+                if not within:
+                    continue
+                tx_index = offset
+                expected_ref = _complement_base(ref.upper())
+                alt_base = _complement_base(alt.upper())
+
+            if 0 <= tx_index < len(seq_list):
+                if expected_ref and seq_list[tx_index].upper() != expected_ref:
+                    continue
+                seq_list[tx_index] = alt_base
+
+        return tx_id, "".join(seq_list)
+
+    # Build mutated canonical transcriptome
+    if build_database:
+        canonical_tx_list = [tx for tx in canonical_tx_ids if tx in transcriptome]
+
+        mutated_canonical_tx_dict = {}
+        if num_threads and num_threads > 1:
+            with concurrent.futures.ThreadPoolExecutor(max_workers=num_threads) as ex:
+                for tx_id, mutseq in ex.map(_apply_snvs_to_one_transcript, canonical_tx_list, chunksize=50):
+                    if mutseq is not None:
+                        mutated_canonical_tx_dict[tx_id] = mutseq
+        else:
+            for tx_id in canonical_tx_list:
+                tx_id2, mutseq = _apply_snvs_to_one_transcript(tx_id)
+                if mutseq is not None:
+                    mutated_canonical_tx_dict[tx_id2] = mutseq
+
+        with open(mutated_canonical_tx_fa, "w") as out:
+            for tx_id, nt_seq in mutated_canonical_tx_dict.items():
+                out.write(f">{tx_id}\n{nt_seq}\n")
+
+        mutated_canonical_transcripts = mutated_canonical_tx_dict
     else:
-        # If you keep your original code, it will write this file when build_database=True.
-        mutated_canonical_transcripts = {}  # placeholder
-    # ---------- END: your existing SNV block should fill these ----------
+        mutated_canonical_transcripts = load_transcriptome(mutated_canonical_tx_fa)
 
-    # Translate mutanome with ATG filter + mapping (even if file was precomputed)
-    mutanome_proteins, mutanome_aa2nt = translate_cds_with_map(mutated_canonical_transcripts, cds_map)
+    # Translate mutanome (CDS only, ATG filtered)
+    mutanome_proteins, mutanome_aa2nt = translate_cds_with_map(
+        mutated_canonical_transcripts,
+        cds_map
+    )
 
     if build_database:
         with open(mutanome_fa, "w") as out:
@@ -1063,8 +1200,7 @@ def classify_peptides(reference,
 
     # -------------------------- alternative ORFs (3 frames) ----------------------
     alt_orf_translatome_fa = os.path.join(database_dir, "mutatedAlternativeTranslatome.fa")
-    alt_orf_orfeome_fa = os.path.join(database_dir, "mutatedAlternativeORFeome.fa")
-    alt_orf_idx_dir = os.path.join(database_dir, "mutatedAlternativeORFeome.idx")
+    alt_orf_idx_dir = os.path.join(database_dir, "mutatedAlternativeTranslatome.idx")
 
     if build_database:
         alt_orf_records = {}
@@ -1082,14 +1218,11 @@ def classify_peptides(reference,
         with open(alt_orf_translatome_fa, "w") as out:
             for rid, aa_seq in alt_orf_records.items():
                 out.write(f">{rid}\n{aa_seq}\n")
-        with open(alt_orf_orfeome_fa, "w") as out:
-            for rid, aa_seq in alt_orf_records.items():
-                out.write(f">{rid}\n{aa_seq}\n")
 
-    # Build index for ORFeome (no aa2nt needed)
+    # Build index for alt ORF translatome (no aa2nt needed)
     if build_fast_index:
         build_proteome_index(
-            alt_orf_orfeome_fa,
+            alt_orf_translatome_fa,
             alt_orf_idx_dir,
             L_min=index_L_min,
             L_max=index_L_max,
@@ -1106,13 +1239,13 @@ def classify_peptides(reference,
             frame = int(frame_str)
         except ValueError:
             return None, None, None, None
-    
+
         nt0 = frame + aa_pos * 3
         if tx_id not in transcriptome:
             return None, None, None, None
         if nt0 < 0 or nt0 >= len(transcriptome[tx_id]):
             return None, None, None, None
-    
+
         if tx_id not in cds_bounds:
             region = "lncRNA"
         else:
@@ -1122,7 +1255,6 @@ def classify_peptides(reference,
             elif nt0 >= cds_end:
                 region = "dORF"
             else:
-                # inside CDS bounds: decide by frame
                 if ((nt0 - cds_start) % 3) == 0:
                     region = "CDS"
                 else:
@@ -1147,7 +1279,7 @@ def classify_peptides(reference,
     # -------------------------- amino acid mismatch ------------------------------
     mismatch_hit_records, peptides_remaining = classify_with_index(
         peptides_remaining,
-        alt_orf_orfeome_fa,
+        alt_orf_translatome_fa,
         alt_orf_idx_dir,
         coord_resolver=coord_resolver_altorf,
         step_label="amino acid mismatch search",

{darkprofiler-0.2.2 → darkprofiler-0.2.3/src/darkprofiler.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: darkprofiler
-Version: 0.2.2
+Version: 0.2.3
 Summary: DarkProfiler: Alignment and Classification of Peptides from Reference-Independent De Novo Peptide Sequencing Experiments.
 Author-email: Hanjun Lee <hanjun@alum.mit.edu>
 License: MIT
@@ -244,7 +244,6 @@ The database contains translated and derived proteomes as FASTA files:
 - `mutanome.fa`
 - `mutatedCanonicalTranscriptome.fa`
 - `mutatedAlternativeTranslatome.fa`
-- `mutatedAlternativeORFeome.fa`
 
 DarkProfiler also creates **persistent fast indices** under the same database directory to accelerate peptide search with Hamming distance:
 for example:
@@ -361,7 +360,7 @@ classify_peptides(
 - **ORF region labels**  
   For alternative ORF hits, DarkProfiler labels the peptide start as:
   - `uORF` (upstream of CDS start)
-  - `intORF` (inside annotated CDS span)
+  - `intORF` (out-of-frame peptdies from inside annotated CDS span)
   - `dORF` (downstream of CDS end)
   - `lncRNA` (no CDS annotation)