build the SUGP1 KD signature with splicedice

2025-01-21T00:00:00+00:00

background, see [[run splicedice on SUGP1 KD signature with splicedice]] for splicedice steps through “pairwise”

raksha called these “reference” SRR12801019 SRR12801023

she called these “sample” SRR12801020 SRR12801024

connect to server

log into prism vpn ssh ubuntu@10.50.100.135

cd /mnt/code/dennisrm_splicedice
git clone https://github.com/dennisrm/splicedice.git
sig=/mnt/code/dennisrm_splicedice/splicedice/code/signature.py

cd /mnt/code/rakshya_RBPSig
git clone https://github.com/codewithrakshya/RBPSig

cd /mnt/code/rakshya_RBPSig/RBPSig
python ./filter_data.py --output-prefix /mnt/output/splicedice_2024.12.19_23.40.49/

iterating

python3 ./filter_data.py --output-prefix /mnt/output/splicedice_2024.12.19_23.40.49/
pip3 install pandas
sudo apt install python3-pandas
sudo apt install python3-argparse
python3 ./filter_data.py --output-prefix /mnt/output/splicedice_2024.12.19_23.40.49/

python3 ./filter_data.hcb.py --output-prefix /mnt/output/splicedice_2024.12.19_23.40.49/

errors addressed

Traceback (most recent call last):
  File "/mnt/code/rakshya_RBPSig/RBPSig/./filter_data.py", line 10, in 
    import pandas as pd
ModuleNotFoundError: No module named 'pandas'

ubuntu@hbeale-mesa:/mnt/code/rakshya_RBPSig/RBPSig$ pip3 install pandas
error: externally-managed-environment

× This environment is externally managed
╰─> To install Python packages system-wide, try apt install
    python3-xyz, where xyz is the package you are trying to
    install.

 cp ./filter_data.py ./filter_data.hcb.py
nano ./filter_data.hcb.py
# changed _pairwiseFisherResults.tsv to pairwise.tsv

copied ./filter_data.py to ./filter_data.hcb.py changed pairwiseFisher to pairwise

ran without error:

ubuntu@hbeale-mesa:/mnt/code/rakshya_RBPSig/RBPSig$ python3 ./filter_data.hcb.py --output-prefix /mnt/output/splicedice_2024.12.19_23.40.49/
Filtered data saved to /mnt/output/splicedice_2024.12.19_23.40.49/_filteredPairwiseResults.tsv
Filtered _allPS data saved to /mnt/output/splicedice_2024.12.19_23.40.49/_filteredAllPS.tsv

signature.py

attempt 1

cd /mnt/output/splicedice_2024.12.19_23.40.49
python3 /mnt/code/dennisrm_splicedice/splicedice/code/signature.py \
fit_beta \
-x significance_threshold=0.05 \
-p /mnt/output/splicedice_2024.12.19_23.40.49/_filteredAllPS.tsv \
-n 10 \
-m  /mnt/output/splicedice_2024.12.19_23.40.49/_manifest.txt \
-o test

ModuleNotFoundError: No module named 'statsmodels'

sudo apt install python3-statsmodels

consider also numpy scipy.stats

tools warnings config

attempt 2

cd /mnt/output/splicedice_2024.12.19_23.40.49
python3 /mnt/code/dennisrm_splicedice/splicedice/code/signature.py \
fit_beta \
-x significance_threshold=0.05 \
-p /mnt/output/splicedice_2024.12.19_23.40.49/_filteredAllPS.tsv \
-n 10 \
-m  /mnt/output/splicedice_2024.12.19_23.40.49/_manifest.txt \
-o test

error when no sig intervals

ubuntu@hbeale-mesa:/mnt/output/splicedice_2024.12.19_23.40.49$ python3 /mnt/code/dennisrm_splicedice/splicedice/code/signature.py fit_beta -x significance_threshold=0.05 -p /mnt/output/splicedice_2024.12.19_23.40.49/_filteredAllPS.tsv -n 10 -m  /mnt/output/splicedice_2024.12.19_23.40.49/_manifest.txt -o test
significance_threshold=0.05
try
Testing for differential splicing...
Groups: /mnt/output/splicedice_2024.12.19_23.40.49/_junction_beds/SRR12801019_Aligned.sortedByCoord.out.junc.bed (1), /mnt/output/splicedice_2024.12.19_23.40.49/_junction_beds/SRR12801020_Aligned.sortedByCoord.out.junc.bed (1), /mnt/output/splicedice_2024.12.19_23.40.49/_junction_beds/SRR12801023_Aligned.sortedByCoord.out.junc.bed (1), /mnt/output/splicedice_2024.12.19_23.40.49/_junction_beds/SRR12801024_Aligned.sortedByCoord.out.junc.bed (1)
Fitting beta distributions...
Traceback (most recent call last):
  File "/mnt/code/dennisrm_splicedice/splicedice/code/signature.py", line 425, in 
    main()
  File "/mnt/code/dennisrm_splicedice/splicedice/code/signature.py", line 87, in main
    beta_stats = manifest.fit_betas(ps_table,compare_stats)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/mnt/code/dennisrm_splicedice/splicedice/code/signature.py", line 328, in fit_betas
    interval_set = self.significant_intervals(compare_stats)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/mnt/code/dennisrm_splicedice/splicedice/code/signature.py", line 303, in significant_intervals
    for i in range(m): # error when no sig intervals
                   ^
UnboundLocalError: cannot access local variable 'm' where it is not associated with a value
ubuntu@hbeale-mesa:/mnt/output/splicedice_2024.12.19_23.40.49$

review input

/mnt/output/splicedice_2024.12.19_23.40.49/_filteredAllPS.tsv - looks fine; there are differences

16:16317777-16317895:+  0.726   0.432   0.736   0.408
16:16317777-16318922:+  0.024   0.145   0.011   0.16
16:16317777-16358956:+  0.0     0.003   0.0     0.003

fix manifest

i think the problem is the manifest. it should be a new one something like this SRR12801019 Reference SRR12801023 Reference SRR12801020 Sample SRR12801024 Sample

 help="TSV file with list of samples (first column) and group labels (second column).")

A19 A20 A23 A24

cd /mnt/output/splicedice_2024.12.19_23.40.49
cp /mnt/code/rakshya_RBPSig/RBPSig/manifest.txt signature_manifest.txt
nano signature_manifest.txt

A19 control
A23 control
A20 SUGP1_knockdown A24 SUGP1_knockdown

attempt 3

cd /mnt/output/splicedice_2024.12.19_23.40.49
python3 /mnt/code/dennisrm_splicedice/splicedice/code/signature.py \
fit_beta \
-x significance_threshold=0.05 \
-p /mnt/output/splicedice_2024.12.19_23.40.49/_filteredAllPS.tsv \
-n 10 \
-m  /mnt/output/splicedice_2024.12.19_23.40.49/signature_manifest.txt \
-o test

ubuntu@hbeale-mesa:/mnt/output/splicedice_2024.12.19_23.40.49$ cd /mnt/output/splicedice_2024.12.19_23.40.49
python3 /mnt/code/dennisrm_splicedice/splicedice/code/signature.py \
fit_beta \
-x significance_threshold=0.05 \
-p /mnt/output/splicedice_2024.12.19_23.40.49/_filteredAllPS.tsv \
-n 10 \
-m  /mnt/output/splicedice_2024.12.19_23.40.49/signature_manifest.txt \
-o test
significance_threshold=0.05
try
Testing for differential splicing...
Groups: control (2), SUGP1_knockdown (2)
Fitting beta distributions...
Traceback (most recent call last):
  File "/mnt/code/dennisrm_splicedice/splicedice/code/signature.py", line 425, in 
    main()
  File "/mnt/code/dennisrm_splicedice/splicedice/code/signature.py", line 87, in main
    beta_stats = manifest.fit_betas(ps_table,compare_stats)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/mnt/code/dennisrm_splicedice/splicedice/code/signature.py", line 328, in fit_betas
    interval_set = self.significant_intervals(compare_stats)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/mnt/code/dennisrm_splicedice/splicedice/code/signature.py", line 303, in significant_intervals
    for i in range(m): # error when no sig intervals
                   ^
UnboundLocalError: cannot access local variable 'm' where it is not associated with a value
ubuntu@hbeale-mesa:/mnt/output/splicedice_2024.12.19_23.40.49$

A Moment on the Wonen’s Rugby Bandwagon

2025-01-18T00:00:00+00:00

I just watched the Bristol Bears play the Leicester Tigers in Premiership Rugby, England’s top division of rugby. I’m still getting the hang of the game, but it was a great game.

I looked at the Bristol Bears roster to remind myself who the player of the game was (Holly Phillips), and noticed that they have three Hollys on the team (OK, one is a Hollie). As a Holly born in the Holly heydey, that’s really surprising. Sure, I was on a sports team with three other Hollys, but there are only 20% as many Hollys now as there were back them.

Documentation