PSP - MetaPredict predicts endogenous disordered regions (IDRs) of protein sequences

Welcome to my CSDN: https://spike.blog.csdn.net/
This article address: https://blog.csdn.net/caroline_wendy/article/details/131461900

Introduction to the MetaPredict algorithm:

Intrinsically Disordered Regions (IDRs) are ubiquitous in all life domains and play various functional roles. While folded domains are often well described by a three-dimensional structure, IDRs exist in a series of interconverted states, called aggregates. This structural heterogeneity means that IDRs are largely missing in the PDB, leading to a lack of computational methods for predicting conformational properties of aggregates from sequences. Here, we combine rational sequence design, large-scale molecular simulations, and deep learning to develop ALBATROSS, a deep learning model for predicting IDR aggregate size from sequences. ALBATROSS enables instantaneous prediction of proteome-wide aggregate-average properties. ALBATROSS is lightweight and easy to use, and is available as a locally installed software package or as a point-and-click cloud interface. We first demonstrate the applicability of our predictor by examining the generalization of sequence-ensemble relations in IDRs. We then exploited the high-throughput nature of ALBATROSS to characterize the emerging biophysical behavior of IDRs both within and beyond the proteome.

Use tools metapredict:

• GitHub：metapredict: A machine learning-based tool for predicting protein disorder.

• Usage documentation: https://metapredict.readthedocs.io/en/latest/

Paper：Direct prediction of intrinsically disordered protein conformational properties from sequence

• Update time 2023-05-28

Online interface: https://metapredict.net/

The reference test document comes from @惠望: https://wiki.biomap-int.com/pages/viewpage.action?pageId=100631778

1. Project configuration

Test T1157s1_A1029.fasta, from CASP15:

>A
DRVRALRRETVEMFYYGFDNYMKVAFPEDELRPVSCTPLTRDLKNPRNFELNDVLGNYSLTLIDSLSTLAILASAPAEDSGTGPKALRDFQDGVAALVEQYGDGRPGPSGVGRRARGFDLDSKVQVFETVIRGVGGLLSAHLFAIGALPITGYQPLRQEDDLFNPPPIPWPNGFTYDGQLLRLALDLAQRLLPAFYTKTGLPYPRVNLRHGIPFYVNSPLHEDPPAKGTTEGPPEITETCSAGAGSLVLEFTVLSRLTGDPRFEQAAKRAFWAVWYRKSQIGLIGAGVDAEQGHWIGTYSVIGAGADSFFEYALKSHILLSGHALPNQTHPSPLHKDVNWMDPNTLFEPLSDAENSAESFLEAWHHAHAAIKRHLYSEREHPHYDNVNLWTGSLVSHWVDSLGAYYSGLLVLAGEVDEAIETNLLYAAIWTRYAALPERWSLREKTVEGGLGWWPLRPEFIESTYHLYRATKDPWYLYVGEMVLRDITRRCWTPCGWAGLQNVLSGEKSDRMESFFLGETTKYMYLLFDDDHPLNKLDASFVFTTEGHPLILPKPKSARRSRNSPRSSQKALTVYQGEGFTNSCPPRPSITPLSGSVIAARDDIYHPARMVDLHLLTTSKHALDGGQMSGQHMAKSNYTLYPWTLPPELLPSNGTCAKVYQPHEVTLEFASNTQQVLGGSAFNFMLSGQNLERLSTDRIRVLSLSGLKITLQLVEEGEREWRVTKLNGIPLGRDEYVVINRAILGDVSDPRFNLVRDPVIAKLQQLHQVNLLDDTTTEEHPDNLDTLDTASAIDLPQDQSSDSEVPDPANLSALLPDLSSFVKSLFARLSNLTSPSPDPSSNLPLNVVINQTAILPTGIGAAPLPPAASNSPSGAPIPVFGPVPESLFPWKTIYAAGEACAGPLPDSAPRENQVILIRRGGCSFSDKLANIPAFTPSEESLQLVVVVSDDEHEGQSGLVRPLLDEIQHTPGGMPRRHPIAMVMVGGGETVYQQLSVASAIGIQRRYYIESSGVKVKNIIVDDGDGGVDG

Install the Python package:

 pip install metapredict==2.61

import metapredict as meta

2. Function call

2.1 Core function Predict Disorder Batch

Test, calculate the probability value of Residue Disorder from 0 to 1, the larger the value, the more likely it is a Disorder site, the threshold is 0.5, and it is mapped to 0 and 1 for binarization, 1 indicates disorder, and 0 indicates fold.

def predict_batch(seq_list):
    if not isinstance(seq_list, list):
        seq_list = [seq_list]

    output = meta.predict_disorder_batch(seq_list)
    assert len(seq_list) == len(output)
    res_list = []
    for sample in output:
        sample_disorder = sample[1]
        print(f"disorder range: {
      
      np.min(sample_disorder)}~{
      
      np.max(sample_disorder)}")
        sample_disorder_idx = list(np.where(sample_disorder > 0.5, 1, 0))
        print(f"sample_disorder_idx: {
      
      sample_disorder_idx}")

        # 获取 disorder 区间
        d_list, tmp_list = [], []
        for i, v in enumerate(sample_disorder_idx):
            if v == 1:  # 无序
                tmp_list.append(i)
            else:
                if tmp_list:
                    d_list.append(copy.copy(tmp_list))
                    tmp_list = []
        domain = []
        for r in d_list:
            if (r[-1] - r[0]) >= 2:
                domain.append([r[0], r[-1]])
        # seq, disorder_idx, domain
        res_list.append([sample[0], sample_disorder_idx, domain])
    return res_list

Among them, the threshold value of 0.5 comes from the suggested value of the algorithm.

Altering the disorder theshhold - To alter the disorder threshold, simply set disorder_threshold=my_value where my_value is a float. The higher the threshold value, the more conservative metapredict will be for designating a region as disordered. Default = 0.5 (V2) and 0.42 (legacy).

output:

disorder range: 0.0~0.9868000149726868
seq:
DRVRALRRETVEMFYYGFDNYMKVAFPEDELRPVSCTPLTRDLKNPRNFELNDVLGNYSLTLIDSLSTLAILASAPAEDSGTGPKALRDFQDGVAALVEQYGDGRPGPSGVGRRARGFDLDSKVQVFETVIRGVGGLLSAHLFAIGALPITGYQPLRQEDDLFNPPPIPWPNGFTYDGQLLRLALDLAQRLLPAFYTKTGLPYPRVNLRHGIPFYVNSPLHEDPPAKGTTEGPPEITETCSAGAGSLVLEFTVLSRLTGDPRFEQAAKRAFWAVWYRKSQIGLIGAGVDAEQGHWIGTYSVIGAGADSFFEYALKSHILLSGHALPNQTHPSPLHKDVNWMDPNTLFEPLSDAENSAESFLEAWHHAHAAIKRHLYSEREHPHYDNVNLWTGSLVSHWVDSLGAYYSGLLVLAGEVDEAIETNLLYAAIWTRYAALPERWSLREKTVEGGLGWWPLRPEFIESTYHLYRATKDPWYLYVGEMVLRDITRRCWTPCGWAGLQNVLSGEKSDRMESFFLGETTKYMYLLFDDDHPLNKLDASFVFTTEGHPLILPKPKSARRSRNSPRSSQKALTVYQGEGFTNSCPPRPSITPLSGSVIAARDDIYHPARMVDLHLLTTSKHALDGGQMSGQHMAKSNYTLYPWTLPPELLPSNGTCAKVYQPHEVTLEFASNTQQVLGGSAFNFMLSGQNLERLSTDRIRVLSLSGLKITLQLVEEGEREWRVTKLNGIPLGRDEYVVINRAILGDVSDPRFNLVRDPVIAKLQQLHQVNLLDDTTTEEHPDNLDTLDTASAIDLPQDQSSDSEVPDPANLSALLPDLSSFVKSLFARLSNLTSPSPDPSSNLPLNVVINQTAILPTGIGAAPLPPAASNSPSGAPIPVFGPVPESLFPWKTIYAAGEACAGPLPDSAPRENQVILIRRGGCSFSDKLANIPAFTPSEESLQLVVVVSDDEHEGQSGLVRPLLDEIQHTPGGMPRRHPIAMVMVGGGETVYQQLSVASAIGIQRRYYIESSGVKVKNIIVDDGDGGVDG
disorder list:
[1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1]
disordered domains:
[[224, 230], [553, 570], [772, 810], [832, 880]]

2.2 Disorder Domains (Official)

predict_disorder_domainsThe predicted value of disorder is predict_disorder_batchconsistent with that of , but the conditions for selecting domains are different and the range is larger.

source code:

output = meta.predict_disorder_domains(seq)
print(output)
print(f"disorder: {output.disorder}")
print(f"disordered_domain_boundaries: {output.disordered_domain_boundaries}")
for boundary in output.disordered_domain_boundaries:
    s = boundary[0]
    e = boundary[1]
    print(f"disorder: {np.min(output.disorder[s:e])} ~ {np.max(output.disorder[s:e])}")
print(f"folded_domain_boundaries: {output.folded_domain_boundaries}")
print(f"disordered_domains: {output.disordered_domains}")
print(f"folded_domains: {output.folded_domains}")

output:

DisorderObject for sequence with 1029 residues, 2 IDRs, and 3 folded domains
Available dot variables are:
  .sequence
  .disorder
  .disordered_domain_boundaries
  .folded_domain_boundaries
  .disordered_domains
  .folded_domains

disorder: [0.5593 0.5207 0.4646 ... 0.7331 0.7067 0.6378]
disordered_domain_boundaries: [[553, 571], [772, 880]]
disorder: 0.5241 ~ 0.854
disorder: 0.2163 ~ 0.9868
folded_domain_boundaries: [[0, 553], [571, 772], [880, 1029]]
disordered_domains: ['KPKSARRSRNSPRSSQKA', 'DDTTTEEHPDNLDTLDTASAIDLPQDQSSDSEVPDPANLSALLPDLSSFVKSLFARLSNLTSPSPDPSSNLPLNVVINQTAILPTGIGAAPLPPAASNSPSGAPIPVF']
folded_domains: ['DRVRALRRETVEMFYYGFDNYMKVAFPEDELRPVSCTPLTRDLKNPRNFELNDVLGNYSLTLIDSLSTLAILASAPAEDSGTGPKALRDFQDGVAALVEQYGDGRPGPSGVGRRARGFDLDSKVQVFETVIRGVGGLLSAHLFAIGALPITGYQPLRQEDDLFNPPPIPWPNGFTYDGQLLRLALDLAQRLLPAFYTKTGLPYPRVNLRHGIPFYVNSPLHEDPPAKGTTEGPPEITETCSAGAGSLVLEFTVLSRLTGDPRFEQAAKRAFWAVWYRKSQIGLIGAGVDAEQGHWIGTYSVIGAGADSFFEYALKSHILLSGHALPNQTHPSPLHKDVNWMDPNTLFEPLSDAENSAESFLEAWHHAHAAIKRHLYSEREHPHYDNVNLWTGSLVSHWVDSLGAYYSGLLVLAGEVDEAIETNLLYAAIWTRYAALPERWSLREKTVEGGLGWWPLRPEFIESTYHLYRATKDPWYLYVGEMVLRDITRRCWTPCGWAGLQNVLSGEKSDRMESFFLGETTKYMYLLFDDDHPLNKLDASFVFTTEGHPLILP', 'LTVYQGEGFTNSCPPRPSITPLSGSVIAARDDIYHPARMVDLHLLTTSKHALDGGQMSGQHMAKSNYTLYPWTLPPELLPSNGTCAKVYQPHEVTLEFASNTQQVLGGSAFNFMLSGQNLERLSTDRIRVLSLSGLKITLQLVEEGEREWRVTKLNGIPLGRDEYVVINRAILGDVSDPRFNLVRDPVIAKLQQLHQVNLL', 'GPVPESLFPWKTIYAAGEACAGPLPDSAPRENQVILIRRGGCSFSDKLANIPAFTPSEESLQLVVVVSDDEHEGQSGLVRPLLDEIQHTPGGMPRRHPIAMVMVGGGETVYQQLSVASAIGIQRRYYIESSGVKVKNIIVDDGDGGVDG']

2.3 Predict pLDDT

metapredictSupports predicting the pLDDT value of Residue, which is used to evaluate the quality of the sequence.

source code:

output = meta.predict_pLDDT(seq)
print(f"mean_plddt: {np.mean(output)}")

2.4 Drawing Graph Disorder

Call directly graph_disorderto draw:

meta.graph_disorder(seq, pLDDT_scores=True)

Disorder Scores are negatively correlated with pLDDT, the plot is as follows:

3. Test structure

Using ChimeraX to test the structure, the command script is as follows:

@classmethod
def get_chimerax_select_cmd(cls, seq_list, mod_num="1"):
    """
    [kaɪˈmɪrə]
    select #1:553-571 #1:772-880
    """
    res_list = cls.predict_batch(seq_list)
    r_str_list = []
    for res in res_list:
        domains = res[2]
        c_list = ["select"]
        for domain in domains:
            c_str = f"#{
      
      mod_num}:{
      
      domain[0]}-{
      
      domain[1]}"
            c_list.append(c_str)
        r_str = " ".join(c_list)
        r_str_list.append(r_str)
    return r_str_list

output:

ChimeraX: select #1:224-230 #1:553-570 #1:772-810 #1:832-880

Test the structure, the green highlight is the disordered area:

other

reference:

• Paper - Direct Prediction of Intrinsically Disordered Protein Conformational Properties From Sequence

• metapredict online (v2.3)

• chimerax-example-scripts-commands

About ChatGPT's translation prompt:

您好，如果你准备好了，请回答“是”，以下是一段蛋白质相关的段落，你作为一个生物学家，帮忙翻译成中文。"[需要翻译的段落]"

The source code is as follows:

#!/usr/bin/env python
# -- coding: utf-8 --
"""
Copyright (c) 2022. All rights reserved.
Created by C. L. Wang on 2023/6/29
"""
import copy
import os

import metapredict as meta
import numpy as np

from protein_utils.seq_utils import get_seq_from_fasta
from root_dir import DATA_DIR


class SeqIdrPredictor(object):
    """
    序列的 IDRs 区域预测
    pip install metapredict==2.61
    """
    def __init__(self):
        pass

    @staticmethod
    def predict_batch(seq_list):
        """
        核心函数
        """
        if not isinstance(seq_list, list):
            seq_list = [seq_list]

        output = meta.predict_disorder_batch(seq_list)
        assert len(seq_list) == len(output)
        res_list = []
        for sample in output:
            sample_disorder = sample[1]
            print(f"disorder range: {
      
      np.min(sample_disorder)}~{
      
      np.max(sample_disorder)}")
            sample_disorder_idx = list(np.where(sample_disorder > 0.5, 1, 0))

            # 获取 disorder 区间
            d_list, tmp_list = [], []
            for i, v in enumerate(sample_disorder_idx):
                if v == 1:  # 无序
                    tmp_list.append(i)
                else:
                    if tmp_list:
                        d_list.append(copy.copy(tmp_list))
                        tmp_list = []
            domains = []
            for r in d_list:
                if (r[-1] - r[0]) >= 2:
                    domains.append([r[0], r[-1]])
            # seq, disorder_idx, domain
            res_list.append([sample[0], sample_disorder_idx, domains])
        return res_list

    @classmethod
    def get_chimerax_select_cmd(cls, seq_list, mod_num="1"):
        """
        [kaɪˈmɪrə]
        select #1:553-571 #1:772-880
        """
        res_list = cls.predict_batch(seq_list)
        r_str_list = []
        for res in res_list:
            domains = res[2]
            c_list = ["select"]
            for domain in domains:
                c_str = f"#{
      
      mod_num}:{
      
      domain[0]}-{
      
      domain[1]}"
                c_list.append(c_str)
            r_str = " ".join(c_list)
            r_str_list.append(r_str)
        return r_str_list

    @staticmethod
    def predict_disorder_domains(seq, is_print=False):
        output = meta.predict_disorder_domains(seq)
        if is_print:
            output = meta.predict_disorder_domains(seq)
            print(output)
            print(f"disorder: {
      
      output.disorder}")
            print(f"disordered_domain_boundaries: {
      
      output.disordered_domain_boundaries}")
            for boundary in output.disordered_domain_boundaries:
                s = boundary[0]
                e = boundary[1]
                print(f"disorder: {
      
      np.min(output.disorder[s:e])} ~ {
      
      np.max(output.disorder[s:e])}")
            print(f"folded_domain_boundaries: {
      
      output.folded_domain_boundaries}")
            print(f"disordered_domains: {
      
      output.disordered_domains}")
            print(f"folded_domains: {
      
      output.folded_domains}")
        return output


def main():
    fasta_dir = os.path.join(DATA_DIR, "CASP15-Monomer-Targets-56", "fasta")
    fasta_path = os.path.join(fasta_dir, "T1157s1_A1029.fasta")
    seq = get_seq_from_fasta(fasta_path)[0]
    sip = SeqIdrPredictor()
    res_list = sip.predict_batch(seq)
    print(f"seq:\n{
      
      res_list[0][0]}")
    print(f"disorder list:\n{
      
      res_list[0][1]}")
    print(f"disordered domains:\n{
      
      res_list[0][2]}")

    sip.predict_disorder_domains(seq, is_print=True)

    output = meta.predict_pLDDT(seq)
    print(f"mean_plddt: {
      
      np.mean(output)}")
    meta.graph_disorder(seq, pLDDT_scores=True)

    r_str_list = sip.get_chimerax_select_cmd(seq)
    print(f"ChimeraX: {
      
      r_str_list[0]}")


if __name__ == '__main__':
    main()