similitudAllProteins.py

import pandas as pd
import Levenshtein
from minineedle import needle, smith, core
from descarteProteinas import substitute_or_remove_prot_id 
from ast import literal_eval
def readData(archivoEntrada):
    """
    Read protein sequences from an Excel file.

    Parameters:
    - archivoEntrada: Input Excel file path
    
    Returns:
    - List of protein sequences

    This function reads protein sequences from an Excel file specified by 'archivoEntrada' and extracts the
    'protein_sequence' column from the DataFrame. The sequences are returned as a list.
    
    Example:
    >>> sequences = readData("protein_data.xlsx")
    >>> print(sequences)
    ['MTCG...', 'MCTA...', ...]
    """
    data = pd.read_excel(archivoEntrada)
    #data=substitute_or_remove_prot_id(data,'r')
    sequences = data["protein_sequence"]

    return sequences

def similitudProteinas(sequences):
    """
    Calculate pairwise similarity scores between protein sequences using Levenshtein distance.

    Parameters:
    - sequences: List of protein sequences
    
    Returns:
    - List of lists containing pairwise similarity information:
        - [protein_sequence_1, protein_sequence_2, similarity_score]

    This function takes a list of protein sequences and calculates pairwise similarity scores
    between each pair of protein sequences using Levenshtein distance. The results are returned
    in a list of lists.

    Example:
    >>> sequences = ["MACG", "MACC", "MGCA"]
    >>> result = similitudProteinas(sequences)
    >>> print(result)
    [['MACG', 'MACC', 75.0],
     ['MACG', 'MGCA', 50.0],
     ['MACC', 'MACG', 75.0],
     ['MACC', 'MGCA', 66.67],
     ['MGCA', 'MACG', 50.0],
     ['MGCA', 'MACC', 66.67]]
    """
    output = []
    for row1 in sequences:
        for row2 in sequences:
            if row1 != row2:
                #similarity = abs(smith.SmithWaterman(row1, row2).get_score()-1) / max(len(row1), len(row2))
                #similarity = abs(needle.NeedlemanWunsch(row1, row2).get_score()-1) / (2*max(len(row1), len(row2)))
                similarity = abs(Levenshtein.distance(row1, row2)) / max(len(row1), len(row2))
                output.append([row1, row2, similarity*100])
    return output

def remplazar_sequence_for_ID(output,archivoEntrada,Sal,mode="default"):
    """
    Replace protein sequences with protein IDs using a pre-existing DataFrame.

    Parameters:
    - output: List of lists containing similarity information
    - mode: Replacement mode (default or drug)
    - archivoEntrada: Path to protein information file
    - Sal: Extension for output file

    This function takes a list of lists containing pairwise similarity information, and replaces
    protein sequences with their corresponding protein IDs. The replacement is based on the information
    provided in a pre-existing DataFrame. The updated information is saved to a CSV file.

    Example:
    >>> data = [['MACG', 'MGCA', 75.0], ['MACC', 'MGCA', 66.67]]
    >>> inputFile = "protein_data.xlsx"
    >>> outputExt = "protein"
    >>> remplazar_sequence_for_ID(data,inputFile,OutputExt, mode="default")
    """
    df_b = pd.read_excel(archivoEntrada)
    #df_b=substitute_or_remove_prot_id(df_b,"r")
    # Ordenar de mayor a menor tamaño. Las subcadenas del mismo tamaño se ordenan por orden alfabetico
    #output_ordered = sorted(output, key=lambda x: (-len(x[0]), x[0]))
    
    proteinas_dict = dict(df_b[['protein_sequence', 'protein_id']].values)
    if(mode=="drug"):
       drug_dict=dict(df_b[['protein_sequence','drug_id']].values)
       for item in output:
        protein_sequence1 = item[0]
        protein_sequence2 = item[1]
        res=[]
        [res.append(x) for x in literal_eval(drug_dict[item[0]]) if x not in res and ( x != '[' or x != ']') ] 
        if(len(res) == 1):
          item.append(res[0])
        elif(len(res)>1):
          item.append(res)
        else:
          item.append("")    
        res=[]
        [res.append(x) for x in literal_eval(drug_dict[item[1]]) if x not in res and ( x != '[' or x != ']')] 
        if(len(res) == 1):
          item.append(res[0])
        elif(len(res)>1):
          item.append(res)
        else:
          item.append("")  
        if protein_sequence1 in proteinas_dict and protein_sequence2 in proteinas_dict:
            item[0] = proteinas_dict[protein_sequence1]
            item[1] = proteinas_dict[protein_sequence2]
       df_a=pd.DataFrame(output, columns=['Proteina1', 'Proteina2', 'Similaridad','SimilaridadAA','similaridadAA_2','drug_id_p1','drug_id_p2'])    
    else:
       for item in output:
        protein_sequence1 = item[0]
        protein_sequence2 = item[1]
        if protein_sequence1 in proteinas_dict and protein_sequence2 in proteinas_dict:
            item[0] = proteinas_dict[protein_sequence1]
            item[1] = proteinas_dict[protein_sequence2]



       df_a = pd.DataFrame(output, columns=['Proteina1', 'Proteina2', 'Similaridad','SimilaridadAA','similaridadAA_2'])

    # Guardar el DataFrame actualizado en un archivo CSV
    df_a.to_csv('AllProteins_%Similitud'+Sal+'.csv', index=False)
def similitudMatProteinas(sequences, matrix,matrix2,matriz3):
    """
    Create percentages of pairwise similarity scores between protein sequences based on three similarity matrices.

    Parameters:
    - sequences: List of protein sequences
    - matrix: First similarity matrix
    - matrix2: Second similarity matrix
    - matriz3: Third similarity matrix

    Returns:
    - List of lists containing pairwise similarity information:
        - [protein_sequence_1, protein_sequence_2, similarity_score_matrix1, similarity_score_matrix2, similarity_score_matrix3]

    This function takes a list of protein sequences and three similarity matrices and calculates pairwise similarity scores
    between each pair of protein sequences. The similarity scores are computed using the provided matrices, and the results
    are returned in a list of lists.

    Note: The function assumes that the matrices are square matrices with dimensions matching the length of the 'sequences' list.

    Example:
    >>> sequences = ["MACG", "MACC", "MGCA"]
    >>> matrix1 = [[1.0, 0.8, 0.6], [0.8, 1.0, 0.7], [0.6, 0.7, 1.0]]
    >>> matrix2 = [[0.9, 0.7, 0.5], [0.7, 0.9, 0.6], [0.5, 0.6, 0.9]]
    >>> matrix3 = [[0.8, 0.6, 0.4], [0.6, 0.8, 0.5], [0.4, 0.5, 0.8]]
    >>> result = similitudMatProteinas(sequences, matrix1, matrix2, matrix3)
    >>> print(result)
    [['MACG', 'MACC', 80.0, 70.0, 60.0],
     ['MACG', 'MGCA', 60.0, 50.0, 40.0],
     ['MACC', 'MACG', 80.0, 70.0, 60.0],
     ['MACC', 'MGCA', 70.0, 60.0, 50.0],
     ['MGCA', 'MACG', 60.0, 50.0, 40.0],
     ['MGCA', 'MACC', 70.0, 60.0, 50.0]]
    """
    output = []
    for row1 in range(0,len(sequences)):
        for row2 in range(0,len(sequences)):
            if row1 != row2:
                #similarity = abs(smith.SmithWaterman(row1, row2).get_score()-1) / max(len(row1), len(row2))
                #similarity = abs(needle.NeedlemanWunsch(row1, row2).get_score()-1) / (2*max(len(row1), len(row2)))
                output.append([sequences[row1], sequences[row2], matrix[row1][row2]*100,matrix2[row1][row2]*100,matriz3[row1][row2]*100])
    return output
if __name__ == "__main__":
    archivoEntrada = "Data/data_nervous_genes_xf.xlsx"
    sequences = readData(archivoEntrada)
    
    matrix=pd.read_csv('matrizNWc.csv',header=None,index_col=False)*3+1
    matrix.abs()
    matrix=matrix/4
    matrix2=pd.read_csv('matrizNWmod1.csv',header=None,index_col=False)*3+1.0
    matrix2.abs()
    matrix2=matrix2/4
    matrix3=pd.read_csv('matrizNWmod2.csv',header=None,index_col=False)*3+1.0
    matrix3.abs()
    matrix3=matrix3/4
    #output = similitudProteinas(sequences)
    output=similitudMatProteinas(sequences, matrix,matrix2,matrix3)
    print("Generada la tabla de con las matrices de similaridad especificadas") 
    remplazar_sequence_for_ID(output,archivoEntrada,"Desease")