Error while running lncDC-train.py

[Jadziaa]

I am trying to run lncDC-train.py and I am getting following error:

File "/home2/jadzia/MARCELA_MOSQUITOS/Ovaries/lncRNA_20_05_24/FEElnc/LncDC-1.3.5/bin/lncDC-train.py", line 375, in
main()
File "/home2/jadzia/MARCELA_MOSQUITOS/Ovaries/lncRNA_20_05_24/FEElnc/LncDC-1.3.5/bin/lncDC-train.py", line 178, in main
mrna_dataset.loc[i, 'CDS_seq'] = cds_data[i]
IndexError: list index out of range

Here is my cmd:

python LncDC-1.3.5/bin/lncDC-train.py -m mrna_AA.fasta -c cds_AA.fasta -l ncRNA.fasta -o lncDC-train_output.txt -t 12 -r

Do you have any recommendations or tips how to solve it?

Thanks a lot!

[lim74]

Hi Monika,

There are probably some compatible issues with the dependencies such as numpy and scikit-learn. I will update the lncDC program with the latest versions of the dependencies and let you know as soon as possible when I'm done!

For your issues, I suggest:

1. Install python and dependencies with the specific versions: python==3.9, numpy==1.23, pandas==1.5, imbalanced-learn==0.9.1 and biopython==1.79. This should solve the compatible issues.
   You can use conda to install a specific version of the dependencies, such as conda install numpy==1.23

2. Are the number of sequences the same between your mrna_AA.fasta and cds_AA.fasta files? I saw that there is an index error when reading the mrna_dataset and cds_data dataframes. It's probably because the numbers of mRNAs and CDSs are different, and when the program loops the mRNA dataframe it cannot find the corresponding CDS of each mRNA.

Please try step 1 first and then step 2. Let me know if you need any other help! The newer version of lncDC will be updated soon!

Best,
Minghua

[lim74]

Hi Monika,

A new version of LncDC has been released (v1.3.6) and the compatible issues should be resolved. You can download and install this latest version for lncRNA model training and prediction. When you train a model, please make sure mRNAs and their corresponding CDSs have the same number and order. Please let me know whether this new version works for you.

Best,
Minghua

[Jadziaa]

Hello,

Thank you soo much for your quick response. I've realised that my mrna_AA.fasta and cds_AA.fasta files had a different number of sequences in it. Fixing this problem solved the issue and everything worked fine for me.

Thank you very much!

Cheers,

Monika

[lim74]

Great! You are welcome!