Dev

README.md

@@ -57,6 +57,14 @@ Easiest is to check [datasets](datasets) examples to see how the above files loo
 
 ## Changelog<a name="changelog"></a>
 
+**v0.8.1**: released in July 2026
+
+* added `bam_file` column support in `samples.tab` for per-sample BAM paths (subfolder layouts)
+* new `get_bam_path(sample_id)` helper in `splicekit/core/annotation.py` — falls back to `{bam_path}/{sample_id}.bam` when no per-sample path is given
+* `exons.py`, `genes.py`, `anchors.py`: replaced `os.listdir()` BAM discovery with `annotation.samples` list + `get_bam_path()` (enables subfolder BAMs, no dir scan needed)
+* `junctions.py`, `jbrowse2.py`: same `get_bam_path()` adoption
+* default `bam_column = "bam_file"` added to config
+
 **v0.8**: released in October 2025
 
 * removed platform config option (now snakemake submits jobs to the cluster)

splicekit/__init__.py

@@ -62,6 +62,7 @@
 splicekit.core.annotation.samples = set()
 splicekit.core.annotation.comparisons = []
 splicekit.core.annotation.genes = {}
+splicekit.core.annotation.bam_path_map = {}
 splicekit.core.annotation.make_comparisons() # load existing comparisons
 
 def setup():

splicekit/config/__init__.py

@@ -74,4 +74,9 @@ def jbrowse2_config():
 try:
     clip
 except:
-    clip = None
+    clip = None
+
+try:
+    bam_column
+except:
+    bam_column = "bam_file"

splicekit/core/anchors.py

@@ -7,6 +7,7 @@
 import os
 import sys
 import splicekit.config as config
+import splicekit.core.annotation as annotation
 import gzip
 
 def write_anchor_gtf():
@@ -69,7 +70,6 @@ def write_jobs_featureCounts(library_type='single-end', library_strand='NONE'):
 
     for anchor_type in ["donor", "acceptor"]:
         gtf_fname = f"reference/{anchor_type}_anchors.gtf.gz"
-        bam_dir = f"{config.bam_path}" # files inside end with <sample_id>.bam
         out_dir = f'data/sample_{anchor_type}_anchors_data'
         jobs_dir = f'jobs/count_{anchor_type}_anchors'
         logs_dir = f'logs/count_{anchor_type}_anchors'
@@ -131,14 +131,13 @@ def write_jobs_featureCounts(library_type='single-end', library_strand='NONE'):
 gzip -f {out_fname}
         """
 
-        bam_files = [fi for fi in os.listdir(bam_dir) if fi.endswith('.bam')]
-        sample_ids = [fi.replace('.bam', '') for fi in bam_files]
+        sample_ids = annotation.samples
         header_line = '\t'.join(['anchor_id', 'count'])
         fsh = open(f"{jobs_dir}/process.sh", "wt")
         for sample in sample_ids:
             out_fname = f"{out_dir}/sample_{sample}.tab"
             job_fname = f'{jobs_dir}/{anchor_type}_anchors_{sample}.job'
-            sam_fname = f"{bam_dir}/{sample}.bam"
+            sam_fname = annotation.get_bam_path(sample)
             # cluster job
             job_out = job_anchors.format(container=config.container, library_type_insert=library_type_insert, library_strand_insert=library_strand_insert, anchor_type=anchor_type, gtf_fname=gtf_fname, sample_id=sample, sam_fname=sam_fname, out_fname=out_fname, logs_dir=logs_dir, header_line=header_line)
             job_file = open(job_fname, "w")

splicekit/core/annotation.py

@@ -56,12 +56,25 @@ def run(self):
         self.returncode = self.process.returncode
         return output.decode("utf-8"), error.decode("utf-8")
 
+def get_bam_path(sample_id):
+    """Return the BAM file path for a sample.
+
+    If samples.tab contained a bam_file column (config.bam_column), returns
+    that per-sample path, allowing BAM files in arbitrary sub-folders.
+    Otherwise falls back to the flat {bam_path}/{sample_id}.bam convention.
+    """
+    if sample_id in annotation.bam_path_map:
+        return annotation.bam_path_map[sample_id]
+    return os.path.join(config.bam_path, f"{sample_id}.bam")
+
 def make_comparisons():
     if not os.path.exists("samples.tab"):
         return
     annotation.comparisons = []
     annotation.treatments = {}
+    annotation.bam_path_map = {}
     samples = set()
+    bam_column = getattr(config, "bam_column", "bam_file")
     f = open("samples.tab")
     r = f.readline()
     while r.startswith("#"):
@@ -80,6 +93,8 @@ def make_comparisons():
         treatment_id = data[config.treatment_column]
         separates.add(data.get(config.separate_column, ""))
         annotation.treatments.setdefault(treatment_id, []).append((sample_id, treatment_id, data.get(config.separate_column, ""), data.get(config.group_column, "")))
+        if bam_column in header and data.get(bam_column, "").strip():
+            annotation.bam_path_map[sample_id] = data[bam_column].strip()
         r = f.readline()
     f.close()
     for treatment, data in annotation.treatments.items(): # sort by sample ID
@@ -173,7 +188,7 @@ def write_comparisons():
         for rtype, rfile in [("control", control_ids), ("test", test_ids)]:
             bams = []
             for sample_id in rfile:
-                bam_fname = os.path.abspath(os.path.join(f"{splicekit.config.bam_path}", f"{sample_id}.bam"))
+                bam_fname = os.path.abspath(get_bam_path(sample_id))
                 bams.append(bam_fname)
             f_rmats = open(f"results/rmats/{comparison_name}_{rtype}.tab", "wt")
             f_rmats.write(",".join(bams))
@@ -454,7 +469,7 @@ def bam_count():
     while r:
         r = r.replace("\r", "").replace("\n", "").split("\t")
         data = dict(zip(header, r))
-        bam_fname = os.path.join(config.bam_path, data[config.sample_column]+".bam")
+        bam_fname = get_bam_path(data[config.sample_column])
         output = subprocess.check_output(f"samtools view -@ 4 -c {bam_fname}", shell=True)
         count = int(output.decode().split("\n")[0])
         print(f"splicekit | bam read counts: {data[splicekit.config.sample_column]}, bam file {bam_fname}, counts = {count}")

splicekit/core/exons.py

@@ -5,6 +5,7 @@
 import sys
 import splicekit.config as config
 import splicekit.core as core
+import splicekit.core.annotation as annotation
 import gzip
 
 def write_exons_gtf():
@@ -61,7 +62,6 @@ def write_jobs_featureCounts(library_type='single-end', library_strand='NONE'):
     library_strand_insert = {"FIRST_READ_TRANSCRIPTION_STRAND":1, "SINGLE_STRAND":1, "SINGLE_REVERSE":1, "SECOND_READ_TRANSCRIPTION_STRAND":2, "NONE":0}[library_strand]
 
     gtf_fname = f"reference/exons.gtf.gz"
-    bam_dir = f"{config.bam_path}" # files inside end with <sample_id>.bam
     out_dir = f'data/sample_exons_data'
     jobs_dir = f'jobs/count_exons'
     logs_dir = f'logs/count_exons'
@@ -124,14 +124,13 @@ def write_jobs_featureCounts(library_type='single-end', library_strand='NONE'):
 gzip -f {out_fname}
     """
 
-    bam_files = [fi for fi in os.listdir(bam_dir) if fi.endswith('.bam')]
-    sample_ids = [fi.replace('.bam', '') for fi in bam_files]
+    sample_ids = annotation.samples
     header_line = '\t'.join(['exon_id', 'count'])
     fsh = open(f"{jobs_dir}/process.sh", "wt")
     for sample in sample_ids:
         out_fname = f"{out_dir}/sample_{sample}.tab"
         job_fname = f'{jobs_dir}/exons_{sample}.job'
-        sam_fname = f"{bam_dir}/{sample}.bam"
+        sam_fname = annotation.get_bam_path(sample)
         # cluster job
         job_out = job_exons.format(container=config.container, library_type_insert=library_type_insert, library_strand_insert=library_strand_insert, gtf_fname=gtf_fname, sample_id=sample, sam_fname=sam_fname, out_fname=out_fname, logs_dir=logs_dir, header_line=header_line)
         job_file = open(job_fname, "w")

splicekit/core/genes.py

@@ -6,6 +6,7 @@
 import sys
 import splicekit.core as core
 import splicekit.config as config
+import splicekit.core.annotation as annotation
 import gzip
 import re
 
@@ -75,7 +76,6 @@ def write_jobs_featureCounts(library_type='single-end', library_strand='NONE'):
     library_strand_insert = {"FIRST_READ_TRANSCRIPTION_STRAND":1, "SINGLE_STRAND":1, "SINGLE_REVERSE":1, "SECOND_READ_TRANSCRIPTION_STRAND":2, "NONE":0}[library_strand]
 
     gtf_fname = f"reference/genes.gtf.gz"
-    bam_dir = f"{config.bam_path}" # files inside end with <sample_id>.bam
     out_dir = f"data/sample_genes_data"
     jobs_dir = f"jobs/count_genes"
     logs_dir = f"logs/count_genes"
@@ -139,14 +139,13 @@ def write_jobs_featureCounts(library_type='single-end', library_strand='NONE'):
     gzip -f {out_fname}
     """
 
-    bam_files = [fi for fi in os.listdir(bam_dir) if fi.endswith('.bam')]
-    sample_ids = [fi.replace('.bam', '') for fi in bam_files]
+    sample_ids = annotation.samples
     header_line = '\t'.join(['gene_id', 'count'])
     fsh = open(f"{jobs_dir}/process.sh", "wt")
     for sample in sample_ids:
         out_fname = f"{out_dir}/sample_{sample}.tab"
         job_fname = f'{jobs_dir}/genes_{sample}.job'
-        sam_fname = f"{bam_dir}/{sample}.bam"
+        sam_fname = annotation.get_bam_path(sample)
         # cluster job
         job_out = job_genes.format(container=config.container, library_type_insert=library_type_insert, library_strand_insert=library_strand_insert, gtf_fname=gtf_fname, sample_id=sample, sam_fname=sam_fname, out_fname=out_fname, logs_dir=logs_dir, header_line=header_line)
         job_file = open(job_fname, "w")

splicekit/core/jbrowse2.py

@@ -1,6 +1,7 @@
 import os
 import splicekit.config as config
 import splicekit.core as core
+import splicekit.core.annotation as annotation
 import http.server
 import socket
 import socketserver
@@ -40,6 +41,7 @@ def setup():
 try:
     bam_files = [fi for fi in os.listdir(bam_dir) if fi.endswith('.bam')]
 except:
+    bam_files = []
     print(f"{module_desc} no bam files found in provided bam folder")
 genome_fa = config.fasta_path
 gff_fname =  config.gff3_path
@@ -100,10 +102,14 @@ def write_sample_jobs(force_samples):
             os.mkdir(dir)
 
     # create jobs
+    # If samples.tab provides per-sample bam paths, use those; otherwise scan bam_dir.
+    if annotation.bam_path_map:
+        sample_bam_pairs = [(sid + ".bam", annotation.get_bam_path(sid)) for sid in annotation.samples]
+    else:
+        sample_bam_pairs = [(fi, bam_dir + '/' + fi) for fi in bam_files]
     bamCoverage_binSize= 3
     fsh = open(f"jobs/jobs_jbrowse/process.sh", "wt")
-    for sample in bam_files:
-        bam_fname = bam_dir +'/'+ sample
+    for sample, bam_fname in sample_bam_pairs:
         cram_fname = dirs_to_check[2] + '/'+sample.replace('.bam', '.cram')
         bigwig_fname = dirs_to_check[2] +'/'+ sample.replace('.bam', '.bw')
         if (not (os.path.exists(cram_fname) and os.path.exists(bigwig_fname))) or (force_samples == True):

splicekit/core/junctions.py

@@ -135,7 +135,7 @@ def make_jobs():
 
     for sample_id in splicekit.core.annotation.samples:
         core_path = os.path.dirname(splicekit.core.__file__)
-        bam_fname = f"{splicekit.config.bam_path}/{sample_id}.bam"
+        bam_fname = splicekit.core.annotation.get_bam_path(sample_id)
         f = open("jobs/count_junctions/sample_{sample_id}.job".format(sample_id=sample_id), "wt")
         f.write(job_junctions.format(sample_id=sample_id, core_path=core_path, bam_fname=bam_fname, job_name="count_junctions_{sample_id}".format(sample_id=sample_id)))
         f.close()

splicekit/version

@@ -1 +1 @@
-0.8
+0.8.1