About rmats_long.py

[border-info-nt]

Hello,

I've completed the analysis with ESPRESSO and am now planning to proceed with rMATS-long analysis. However, I generated the .esp and updated.gtf files individually for each sample, so I’m unable to perform a multi-sample analysis. Is it acceptable to merge these files, or should I start the analysis from scratch?

For exmple

• transcript_ID transcript_name gene_ID gs689_1 gs689_2 gs689_3 pc3e_1 pc3e_2 pc3e_3
• ENST00000428599.6 CTNND1-208 ENSG00000198561.16 23.66 16.47 30.81 0.60 0.36 0.31
• ENST00000682814.1 CTNND1-245 ENSG00000198561.16 0 0 0 1.00 1.01 0
• ENST00000681984.1 CTNND1-242 ENSG00000198561.16 0 0 1.54 0 0 1.27
• ENST00000683906.1 CTNND1-250 ENSG00000198561.16 5.45 1.21 2.33 1.11 0 1.03
• ENST00000426142.6 CTNND1-207 ENSG00000198561.16 2.89 8.82 4.42 15.44 15.73 6.54
• ENST00000683769.1 CTNND1-249 ENSG00000198561.16 2.13 2.31 0 2.18 0 1.01
• ENST00000532463.5 CTNND1-230 ENSG00000198561.16 5.90 0 7.31 1.75 6.94 6.54
• ENST00000529986.5 CTNND1-222 ENSG00000198561.16 24.91 25.03 59.60 54.96 54.33 65.52
• ENST00000358694.10 CTNND1-201 ENSG00000198561.16 48.22 73.37 111.83 0.60 0.36 0.31
• ENST00000684704.1 CTNND1-252 ENSG00000198561.16 0 0 1.04 0 0 0
• ENST00000530068.5 CTNND1-223 ENSG00000198561.16 0 0 0 1.00 0 0
• ENST00000531007.2 CTNND1-227 ENSG00000198561.16 0 3.23 0 0 0.28 0.29
• • ENST00000534579.5 CTNND1-236 ENSG00000198561.16 5.09 0 5.01 0.60 0.36 0.31

How can I output [gs689_1 gs689_2 gs689_3 pc3e_1 pc3e_2 pc3e_3] in espresso like this? Or should I merge the results output one sample at a time?

[EricKutschera]

Ideally the espresso run would include all of your samples and the .esp from espresso would then have all the sample names as columns. When you run espresso you can list the sample names along with the input files in the samples.tsv file described in the README: https://github.com/Xinglab/espresso/tree/v1.5.0?tab=readme-ov-file#basic-usage

It's not easy to merge the results from multiple espresso runs since the novel isoforms detected would have different IDs (ESPRESSO:chr1:2:3). Also there may be novel isoforms which have slightly different transcript start or end coordinates between runs but are otherwise the same