Merge branch 'pr/100'

# Conflicts:
#	AppImageBuilder.yml
#	web/auto-imports.d.ts
#	web/components.d.ts
#	web/package.json
#	web/pnpm-lock.yaml

Author: ygidtu

docs/command.md

@@ -661,7 +661,63 @@ trackplot \
 
 ![](imgs/cmd/intron_scale.png)
 
-### 7. Visualize coverage by cpm or rpkm
+
+### 7. Fixed intron size
+
+
+There are two BAM files. Here is the output with introns fixed at 250 nt, exons unharmed (1:1), without any filtering of low-count junctions:
+
+![](imgs/cmd/PTBP3_mod_specific_transcipts.png)
+
+With --log 10, we can see all the low-count junctions:
+
+![](imgs/cmd/PTBP3_mod_specific_transcipts_log10_edit.png)
+
+If we use Trackplot's scaling features, we can get reasonably close (exons=20, introns=0.10):
+
+![](imgs/cmd/PTBP3_mod_specific_transcipts_introns_and_exons_scaled.png)
+
+Here is the equivalent plot with Trackplot (note the gaps remaining from excluded transcripts):
+
+![](imgs/cmd/PTBP3_trackplot_specific_trans_all_junctions.exons.20.introns.0.10.png)
+
+With --density-by-strand:
+
+![](imgs/cmd/PTBP3_trackplot_all_junctions_specific_transcripts_by_strand_edit.png)
+
+
+To be clear, the modifications described here were a small step following a giant leap forward. For instance, here is how the data looks in IGV:
+
+![](imgs/cmd/ptbp3_igv_2.png)
+
+
+The switch to positive-only Y values is hard-coded, along with other preferences (transparency etc.).
+
+To replicate the figures above, use:
+
+```bash
+trackplot \
+  --intron-scale 250  \   # To assigned a fixed length to introns, use an intron scale greater than 1
+  -e 9:112225716-112333664 \
+  -r ptbp3_ext.gtf.gz \
+  --show-junction-num \
+  --density bam_files.tsv \
+  --raster \
+  --height 2 \
+  --width 20 \
+  --font-size 14 \
+  --transcripts-to-show PTBP3-201,PTBP3-202,PTBP3-203,PTBP3-204,PTBP3-205,PTBP3-206,PTBP3-207,PTBP3-208 \
+  -o PTBP3_mod_specific_transcipts.log10.png
+```
+
+To scale using exising Trackplot features, use:
+
+```bash
+  --exon-scale 20 \
+  --intron-scale 0.10  \
+```
+
+### 8. Visualize coverage by cpm or rpkm
 
 We also support to visualize the coverage by normalized values.
 
@@ -1119,4 +1175,6 @@ trackplot \
   --line example/line_list.tsv
 ```
 
-![](imgs/cmd/additional.png)
+![](imgs/cmd/additional.png)
+
+

trackplot/file/Annotation.py

@@ -2,8 +2,10 @@
 # -*- coding:utf-8 -*-
 u"""
 Created by ygidtu@gmail.com at 2020.05.07
-
 This scripts contains the class handle the reference file
+
+Modified by AD 2025/01/12 to include only specified transcripts, otherwise their exon coordinates still occupy the alignments.
+
 """
 import glob
 import gzip
@@ -349,16 +351,15 @@ def index_gtf(cls, input_gtf):
 
         return output_gtf
 
-    def __load_gtf__(self):
-
+    def __load_gtf__(self, transcripts_to_show: list[str]|None = None):
         u"""
         Load transcripts inside of region from gtf file
         :param region: target region
         :return: list of Transcript
-        """
+        """ 
+        # AD - passing transcripts_to_show
         transcripts = {}
         exons = {}
-
         for rec in Reader.read_gtf(self.path, self.region):
             start = max(rec.start, self.region.start)
             end = min(rec.end, self.region.end)
@@ -369,6 +370,13 @@ def __load_gtf__(self):
                 break
 
             if re.search(r"(rna|transcript|cds)", rec.feature, re.I):
+
+                if transcripts_to_show:
+                    _name = rec.transcript_name if "transcript_name" in rec.attributes else rec.transcript_id
+                    if _name not in transcripts_to_show:
+                        logger.info(f"Skipping transcript {_name}")
+                        continue
+
                 if rec.transcript_id not in transcripts.keys():
                     transcripts[rec.transcript_id] = Transcript(
                         chromosome=rec.contig,
@@ -380,7 +388,8 @@ def __load_gtf__(self):
                         gene=rec.gene_name if "gene_name" in rec.attributes else "",
                         transcript=rec.transcript_name if "transcript_name" in rec.attributes else "",
                         exons=[]
-                    )
+                 )
+
             elif re.search(r"(exon)", rec.feature, re.I):
                 if rec.transcript_id not in exons.keys():
                     exons[rec.transcript_id] = []
@@ -581,9 +590,16 @@ def load(self,
         assert isinstance(region, GenomicLoci), "region should be a GenomicLoci object"
         if transcripts is None:
             transcripts = []
+        elif isinstance(transcripts, str):
+            # AD when no transcripts are specified, transcripts is an empty string, not None
+            if transcripts == "":
+                transcripts = []
+            else:
+                transcripts = transcripts.split(",")
         self.region = region
+        # AD - pass specified transcripts early
         if self.category == "gtf":
-            self.__load_gtf__()
+            self.__load_gtf__(transcripts)
             if self.add_local_domain:
                 self.__load_local_domain__(region)
 

trackplot/file/ReadSegments.py

@@ -382,7 +382,9 @@ def load_bam(self):
                     current_ignore_num = min(
                         [self.deletion_ignore, read.query_alignment_length * self.del_ratio_ignore])
                 else:
-                    current_ignore_num = np.Inf
+                    current_ignore_num = np.inf
+                    # AD - AttributeError: `np.Inf` was removed in the NumPy 2.0 release. Use `np.inf` instead.
+
 
                 exon_bound = []
                 intron_bound = []

trackplot/plot.py

@@ -1437,7 +1437,7 @@ def plot(self,
 
         if output:
             logger.info(f"saving fig into {output}")
-            fig.savefig(output, transparent=True, bbox_inches='tight')
+            fig.savefig(output, transparent=False, bbox_inches='tight') # AD
         elif return_image:
             output = io.BytesIO()
             if return_image == "png":