PDBRX report: "edit the system segmentation"

[cjeong73]

• Input pdb file, 4F87.pdb has 4 chains(A,B,C, and D) with missing residues in each chain.

• PDBSCAN reported the detailed information of sequences including missing residues. According to the log of PDBSCAN, fasta sequence should be...

A> mskinvnvenVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMKk B> mskinvnveNVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMKk C> mskinvnveNVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMKk D> mskinvnveNVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMKk

• In case of Non-default mode, PDBRX in dev branch generated the sequences as below. Note Chain B,C had no missing residues while chain A, D have missing residue(s) at one chain end. Another issue is that the output structure was made based on this sequence info.

Current sequences (lowercase indicates residues not in coordinates): A: mskinvnvenVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMK C: NVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMK B: NVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMK D: NVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMKk

• PDBRX in jetstream resulted in different sequences.

A: VSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMK C: NVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMK B: NVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMK D: NVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMK

[cjeong73]

In terminal_edit_options module of preprocess.py, I replace mol.segname_info.sequence_to_fasta with mol.chain_info.sequence_to_fasta to get fasta sequence.

        print("Current sequences (lowercase indicates residues not in coordinates): ")

        for segname in seq_segnames:
          #  seq = mol.segname_info.sequence_to_fasta(
            seq = mol.chain_info.sequence_to_fasta(
                segname, missing_lower=True)
            print(segname + ':')
            print(seq)

Then, non-default mode printed out the segmentation info correctly as below. Need to find out the occurrence of this error in the codes.

Current sequences (lowercase indicates residues not in coordinates): A: mskinvnvenVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMKk C: mskinvnveNVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMKk B: mskinvnveNVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMKk D: mskinvnveNVSGVQGFLFHTDGKESYGYRAFINGVEIGIKDIETVQGFQQIIPSINISKSDVEAIRKAMKk

[madscatt]

While this gives the correct output it is only for a "print" statement.

[cjeong73]

Yes, it is only for print statement. Then, I tested what happens if segname_info is replaced with chain_info in all pdbrx codes of dev branch at onsager. This change corrected the output psf and pdb following the sequence info of pdbscan at least for dev branch codes.

[madscatt]

That is a possible solution, perhaps a better solution (as segname/segment is the currency involved) is to update the segname_info object. I am very uneasy with finding solutions by replacing bits of code with out understanding the consequences.

[madscatt]

See issue report in #117

[dww100]

@cjeong73 What is the status of this post the fixes for #117?