Functions.py

from Graph_gff import Segments, Features, get_feature_start_on_segment, get_feature_stop_on_segment
global segments_on_target_genome
segments_on_target_genome={}

# get the start position of the features on the linear genome, using their coordinates on the graph and the coordinantes of the segments on the genome
def get_feature_start_on_genome(start_seg,feat_id):
    seg_start_pos=segments_on_target_genome[start_seg][1]
    feat_start_pos=get_feature_start_on_segment(start_seg,feat_id)
    
    start=int(seg_start_pos)+int(feat_start_pos)-1
    return start

# get the stop position of the features on the linear genome, using their coordinates on the graph and the coordinantes of the segments on the genome
def get_feature_stop_on_genome(stop_seg,feat_id):
    seg_start_pos=segments_on_target_genome[stop_seg][1]
    feat_stop_pos=get_feature_stop_on_segment(stop_seg,feat_id)

    stop=int(seg_start_pos)+int(feat_stop_pos)-1
    return stop



# functions to get the detailed gff with all the fragments of the features

# get the position of a part of a feature on the complete feature (on the original genome)
def get_position_on_feature(start_seg,stop_seg,feature_start_segment,feature):
    start_on_feature=get_segment_start_on_feature(feature_start_segment,start_seg,feature)
    stop_on_feature=get_segment_stop_on_feature(start_seg,stop_seg,feature,start_on_feature)
    position=f';position={start_on_feature}-{stop_on_feature}'
    return position

def get_segment_start_on_feature(feature_start_segment,start_seg,feature):
    feature_start_pos=Segments[feature_start_segment].start+get_feature_start_on_segment(feature_start_segment,feature.id)-1
    start_on_reference=Segments[start_seg].start+get_feature_start_on_segment(start_seg,feature.id)-1
    start_on_feature=int(start_on_reference)-int(feature_start_pos)+1
    return start_on_feature

def get_segment_stop_on_feature(start_seg,stop_seg,feature,start_on_feature):
    start_on_new_genome=get_feature_start_on_genome(start_seg,feature.id)
    stop_on_new_genome=get_feature_stop_on_genome(stop_seg,feature.id)
    # length on new genome is the same as length on reference, as we only get the positions of conserved parts of the feature
    stop_on_feature=start_on_feature+(stop_on_new_genome-start_on_new_genome) # stop position : start+length
    return stop_on_feature

# get the proportion of a part of the feature on the total length
def get_proportion_of_feature_part(start_seg,stop_seg,feature):
    start_on_new_genome=get_feature_start_on_genome(start_seg,feature.id)
    stop_on_new_genome=get_feature_stop_on_genome(stop_seg,feature.id)
    # length on new genome is the same as length on reference, as we only get the positions of conserved parts of the feature
    proportion=f';proportion={stop_on_new_genome-start_on_new_genome+1}/{feature.size}'
    return proportion

# returns the gff line to write in the output file for the function gff_detail
def create_line_detail(last_seg,feature_id,start_seg,feat_start):
    [stop_seg,feature,chr,strand]=[last_seg[1:],Features[feature_id],segments_on_target_genome[start_seg][0],Features[feature_id].strand]
    # start and stop position of the feature on the genome we transfer on
    start_on_new_genome=get_feature_start_on_genome(start_seg,feature_id) 
    stop_on_new_genome=get_feature_stop_on_genome(stop_seg,feature_id)
    
    proportion=get_proportion_of_feature_part(start_seg,stop_seg,feature)
    position=get_position_on_feature(start_seg,stop_seg,feat_start,feature)
    annotation=feature.annot+proportion+position
    
    output_line=f'{chr}\tGrAnnot\t{feature.type}\t{start_on_new_genome}\t{stop_on_new_genome}\t.\t{strand}\t.\t{annotation}\n'
    return output_line


# functions to get the gff with one line per feature
def right_size(size,max_diff,feat):
    if max_diff==0:
        return True
    return not ((size>Features[feat].size*max_diff) | (size<Features[feat].size/max_diff)) 

def create_line_gff_one(first_seg,last_seg,feature_id,size_diff,list_seg):
    [chr,strand,feature]=[segments_on_target_genome[first_seg][0],Features[feature_id].strand,Features[feature_id]]

    variant_count=0
    for segment in list_seg:
        if segment[1:] not in segments_on_target_genome: # if a segment is absent, it is either a deletion of a substitution. insertions are not considered here.
            variant_count+=1
    annotation=f'{feature.annot};Size_diff={size_diff};Nb_variants={variant_count}'

    output_line=f'{chr}\tGrAnnot\t{feature.type}\t{get_feature_start_on_genome(first_seg,feature_id)}\t{get_feature_stop_on_genome(last_seg,feature_id)}\t.\t{strand}\t.\t{annotation}\n'
    return output_line


# functions to output the stats on the transfer

def stats_feature_missing_segment(feature_missing_segments,first_seg,last_seg,list_seg,feature_id):
# [feature_missing_first,feature_missing_middle,feature_missing_last,feature_missing_all,feature_missing_total,feature_total,feature_ok]
    feature_missing_segments[5].append(feature_id)

    if (first_seg=='') : # no segment of the feature is in the genome, the feature is missing entirely
        feature_missing_segments[3].append(feature_id)
    elif first_seg != list_seg[0][1:]: # the first segment is missing 
        feature_missing_segments[0].append(feature_id)
    elif last_seg!=list_seg[-1][1:]: # the last segment is missing
        feature_missing_segments[2].append(feature_id)

    # go through all the segments, check if some are missing in the middle of the feature
    elif (len(list_seg)!=1) & (feature_id not in feature_missing_segments[3]): # to access the second to last element
        for segment in list_seg[1-(len(list_seg)-2)]:
            if segment[1:] not in segments_on_target_genome:
                feature_missing_segments[1].append(feature_id)
                break

    # go through the segments, to see if one is missing anywhere on the feature
    for segment in list_seg:
        if segment[1:] not in segments_on_target_genome:
            if feature_id not in feature_missing_segments[4]:
                feature_missing_segments[4].append(feature_id)
                break

    # if the feature doesnt have a missing segment, it is complete.         ADD THE PATH CHECK FOR INSERTIONS !!
    if feature_id not in feature_missing_segments[4]:
        feature_missing_segments[6].append(feature_id)

def get_annot_features(list_features):
    list_annot_features=[]
    for feature in list_features:
        list_annot_features.append(Features[feature].note)
    return list_annot_features

def count_hypput_total(list_annot_first):
    total=len(list_annot_first)
    count_hypput=0
    for annot in list_annot_first:
        if ("hypothetical" in annot) | ("putative" in annot):
            count_hypput+=1
    return [count_hypput,total]

# print stats on the transfer : number of feature that have segments in different positions missing. 
def stats_features(feature_missing_segments):
# [feature_missing_first,feature_missing_middle,feature_missing_last,feature_missing_all,feature_missing_total,feature_total,feature_ok]
    list_annot_first=get_annot_features(feature_missing_segments[0])
    [hyp_put,total]=count_hypput_total(list_annot_first)
    print("\nthe first segment is missing for", total,"features, including",round(100*(hyp_put)/total,2),"% hypothetical or putative.")

    list_annot_middle=get_annot_features(feature_missing_segments[1])
    [hyp_put,total]=count_hypput_total(list_annot_middle)
    print("a middle segment is missing for", total,"features, including",round(100*(hyp_put)/total,2),"% hypothetical or putative.")

    list_annot_last=get_annot_features(feature_missing_segments[2])
    [hyp_put,total]=count_hypput_total(list_annot_last)
    print("the last segment is missing for", total,"features, including",round(100*(hyp_put)/total,2),"% hypothetical or putative.")

    list_annot_all=get_annot_features(feature_missing_segments[3])
    [hyp_put,total]=count_hypput_total(list_annot_all)
    print(total,"features are entirely missing, including",round(100*(hyp_put)/total,2),"% hypothetical or putative.")

    list_annot_total=get_annot_features(feature_missing_segments[4])
    [hyp_put,total]=count_hypput_total(list_annot_total)
    print("there is at least one segment missing for", total,"features, including",round(100*(hyp_put)/total,2),"% hypothetical or putative.")

    list_annot_ok=get_annot_features(feature_missing_segments[6])
    [hyp_put,total]=count_hypput_total(list_annot_ok)
    print(total ,"features are entirely present in the new genome, including",round(100*(hyp_put)/total,2),"% hypothetical or putative.")

    list_annot_features=get_annot_features(feature_missing_segments[5])
    [hyp_put,total]=count_hypput_total(list_annot_features)
    print("there is", total,"features in total, including",round(100*(hyp_put)/total,2),"% hypothetical or putative.")




# functions to generate the different gffs

def get_segments_positions_on_genome(pos_seg):
    bed=open(pos_seg,'r')
    lines=bed.readlines() # read line by line ?
    bed.close()
    for line in lines:
        line=line.split()
        [seg,chrom,start,stop,strand]=[line[3][1:],line[0],line[1],line[2],line[3][0:1]]
        segments_on_target_genome[seg]=[chrom,start,stop,strand]

def get_segments_sequence_and_paths(gfa):
    file_gfa=open(gfa,'r')
    lines_gfa=file_gfa.readlines()
    file_gfa.close()
    seg_seq={}
    paths={}
    for line in lines_gfa:
        line=line.split()
        if (line[0]=="S"): # get the sequence of the segment
            seg_id='s'+line[1]
            seg_seq[seg_id]=line[2]

        if (line[0]=="W") & (line[1]!="_MINIGRAPH_"): # get the walk of the genome
            
            path=line[6].replace(">",";>")
            path=path.replace("<",";<").split(';')
            list_path=[]
            for segment in path:
                if segment[0:1]=='>':
                    list_path.append('+s'+segment[1:])
                elif segment[0:1]=='<':
                    list_path.append('-s'+segment[1:])
                else:
                    list_path.append('s'+segment[1:])
 
            paths[line[3]]=list_path
    return [paths,seg_seq]

def get_first_seg(list_seg):
    first_seg_found=''
    for segment in list_seg:
        if segment[1:] in segments_on_target_genome:
            first_seg_found=segment[1:]
            break
    return first_seg_found

def get_all_features_in_gff(gff):
    gff=open(gff,'r')
    lines=gff.readlines()
    gff.close()
    list_feature=[]
    for line in lines:
        feature_id=(line.split()[8].split(";")[0].split("=")[1].replace(".","_").replace(":","_"))
        if feature_id not in list_feature:
            list_feature.append(feature_id)
    return list_feature



# functions to get the detail of the variations in the features

def compare_strand(list_1,list_2,list_1_unstrand,list_2_unstrand):
    # get the list of segments in common
    seg_common=[]
    for segment in list_1_unstrand:
        if segment in list_2_unstrand:
            seg_common.append(segment)

    # for each segment in common, check if the strand is the same. check index in list unstranded to get the segment in list stranded
    same_strand_count=0
    for segment in seg_common:
        index_1=list_1_unstrand.index(segment)
        index_2=list_2_unstrand.index(segment)
        if list_1[index_1]==list_2[index_2]:
            same_strand_count+=1
    return [seg_common,same_strand_count]

def get_feature_path(paths,first_seg,last_seg):
    # find the path in azucena. 
    first_strand=convert_strand(segments_on_target_genome[first_seg][3])
    first_seg_stranded=first_strand+first_seg
    last_strand=convert_strand(segments_on_target_genome[last_seg][3])
    last_seg_stranded=last_strand+last_seg
    index_first_seg=int(paths["CM020642.1_Azucena_chromosome10"].index(first_seg_stranded))
    index_last_seg=int(paths["CM020642.1_Azucena_chromosome10"].index(last_seg_stranded))
    first_index=min(index_first_seg,index_last_seg)
    last_index=max(index_last_seg,index_first_seg)
    list_segfeat_azu=paths["CM020642.1_Azucena_chromosome10"][first_index:last_index+1]
    list_segfeat_azu_corrected=[convert_strand(segment_stranded[0])+segment_stranded[1:] for segment_stranded in list_segfeat_azu]
    return list_segfeat_azu_corrected

def convert_strand(strand):
    match strand:
        case "+":
            return ">"
        case "-":
            return "<"
        case ">":
            return "+"
        case "<":
            return "-"
        case default:
            return ""

def get_sequence_list_seg(list_segfeat_nb,i,feature,seg_seq):
    del_sequence=""
    for k in range(i,len(list_segfeat_nb)):
        if k==len(list_segfeat_nb)-1:
            del_sequence+=seg_seq[list_segfeat_nb[k][1:]][0:feature.pos_stop]
        else:
            del_sequence+=seg_seq[list_segfeat_nb[k][1:]]
    return del_sequence

class Variation:
    def __init__(self,feature_id,feature_type,chr,start_new,stop_new,inversion,size_diff):
        self.feature_id=feature_id
        self.feature_type=feature_type
        self.chr=chr
        self.start_new=start_new
        self.stop_new=stop_new
        self.inversion=str(inversion)
        self.size_diff=str(size_diff)
        self.size_new=str(self.stop_new-self.start_new+1)
        self.type=''
        self.last_seg_in_target=''
        
    # ajouter fct pour écrire la ligne. return line.

    #def __str__(self):
    #    return f"id={self.id}, position on the original genome={self.chr}:{self.start}-{self.stop}, size={self.size}, features={self.features}"

def create_var(feature_id,first_seg,last_seg,paths):
    feature=Features[feature_id]
    start_new_genome=get_feature_start_on_genome(first_seg,feature_id)
    stop_new_genome=get_feature_stop_on_genome(last_seg,feature_id)
    size_new_genome=int(stop_new_genome)-int(start_new_genome)+1
    size_diff=str(size_new_genome-feature.size)

    # get feature paths on the original genome and on the target genome
    list_segfeat_azu=get_feature_path(paths,first_seg,last_seg)
    list_segfeat_nb=feature.segments_list
    [list_segfeat_nb,list_segfeat_azu,inversion]=detect_inversion(list_segfeat_nb,list_segfeat_azu)

    variation=Variation(feature_id,feature.type,feature.chr,start_new_genome,stop_new_genome,inversion,size_diff)

    return(variation,list_segfeat_nb,list_segfeat_azu)

def reset_var(variation):
    variation.type='' # type énuméré.
    variation.size_var=0
    variation.start_var=''
    variation.ref=''
    variation.alt=''

def get_old_new_pos_substitution(feat_start,variation,list_segfeat_azu,feat,j):
    pos_old=str(int(Segments[variation.start_var].start)-int(feat_start)+1)
    start_feat=get_feature_start_on_genome(list_segfeat_azu[0][1:],feat)
    start_var=int(segments_on_target_genome[variation.start_on_target][1])
    pos_new=str(start_var-start_feat+1)
    return [pos_old,pos_new]

def get_old_new_pos_insertion(variation,feat_start,list_segfeat_azu,feat):
    pos_old=str(int(Segments[variation.start_var].start)-int(feat_start)+1)

    start_feat=get_feature_start_on_genome(list_segfeat_azu[0][1:],feat) 
    start_var=int(segments_on_target_genome[variation.start_var][1])-1
    pos_new=str(start_var-start_feat+1)
    return [pos_old,pos_new]

def get_old_new_pos_deletion(variation,feat_start,list_segfeat_azu,feat,i):
    if i==0:
        pos_old=int(Segments[variation.start_var].start)-int(feat_start)+1+Features[feat].pos_start
    else:
        pos_old=int(Segments[variation.start_var].start)-int(feat_start)+1
        if pos_old<0:
            pos_old=0

    if variation.last_seg_in_target=="": # deletion of the beggining of the feature, so no segment placed in the new genome yet. 
        pos_new="1"
    else:
        start_feat=get_feature_start_on_genome(list_segfeat_azu[0][1:],feat) 
        start_var=int(segments_on_target_genome[variation.last_seg_in_target][2])+1 
        pos_new=str(start_var-start_feat+1)
    
    return [pos_old,pos_new]

def init_new_var(variation,type,list_segfeat_nb,list_segfeat_azu,i,j,seg_seq,feature):
    variation.type=type
    variation.start_var=list_segfeat_nb[i][1:]
    if type=="substitution":
        variation.start_on_target=list_segfeat_azu[j][1:]
        variation.ref=seg_seq[list_segfeat_nb[i][1:]]
        variation.alt=seg_seq[list_segfeat_azu[j][1:]]
    elif type=="insertion":
        variation.ref="-"
        variation.alt=seg_seq[list_segfeat_azu[j][1:]]
    elif type=="deletion":
        if i==0: # if the deletion is at the start of the feature, the deletion doesnt always start at the start at the first segment : 
            #use pos_start, position of the feature on its first segment
            variation.ref=seg_seq[list_segfeat_nb[i][1:]][feature.pos_start-1:]
        else: # else, the deletion will always start at the start of the first segment.
            variation.ref=seg_seq[list_segfeat_nb[i][1:]]
        variation.alt="-"

def continue_var(variation,seg_seq,list_segfeat_nb,list_segfeat_azu,i,j):

    if variation.type=="substitution":
        variation.ref+=seg_seq[list_segfeat_nb[i][1:]]
        variation.alt+=seg_seq[list_segfeat_azu[j][1:]]
    elif variation.type=="insertion":
        variation.alt+=seg_seq[list_segfeat_azu[j][1:]]
    elif variation.type=="deletion":
        variation.ref+=seg_seq[list_segfeat_nb[i][1:]]

def get_common_segments(list1,list2):
    list_output=[]
    for elem in list1:
        if elem in list2:
            list_output.append(elem)
    return list_output

def detect_segment_order_inversion(list_1,list_2):
    if (len(list_1)==1) | (len(list_2)==1):
        return False
    [cpt,i]=[0,0]
    list_1_common=get_common_segments(list_1,list_2)
    list_2_common=get_common_segments(list_2,list_1)
    list_2_common_reversed=list(reversed(list_2_common))
    while i<len(list_1_common):
        if list_2_common_reversed[i]==list_1_common[i]:
            cpt+=1
        i+=1
    return (cpt>len(list_1_common)*0.9) # if more than 90% of the segments are on the same position when the lists are reversed, there is an inversion. 
    
def detect_orient_inversion(list_1,list_2):
    list_1_unstrand=[segment_stranded[1:] for segment_stranded in list_1]
    list_2_unstrand=[segment_stranded[1:] for segment_stranded in list_2]
    [seg_common,same_strand_count]=compare_strand(list_1,list_2,list_1_unstrand,list_2_unstrand)

    if same_strand_count>=len(seg_common)*0.9: # if more than 90% of segments shared have the same strand, no inversion
        strand_inversion=False
    else:
        strand_inversion=True
        if same_strand_count!=0:
            print("ignored strand difference :/")
    return [strand_inversion,list_1_unstrand,list_2_unstrand]


# takes two lists of segments for two genes, check if the first list is an inversion of the second one (if the segments in common are on the opposite strand)
def detect_inversion(list_1,list_2):
    # check if we have an inversion of the orientation of the segments
    [strand_inversion,list_1_unstrand,list_2_unstrand]=detect_orient_inversion(list_1,list_2)

    # check if we have an inversion of the order of the segments
    segment_order_inversion=detect_segment_order_inversion(list_1_unstrand,list_2_unstrand)

    # if there we have both inversions, the gene is in an inverted region. reverse the second list for the comparison.
    if segment_order_inversion & strand_inversion:
        inversion=1
        list_2=invert_segment_list(list_2)
    else :
        inversion=0

    return [list_1,list_2,inversion]

def invert_segment_list(seg_list):
    list_inverted=list()
    for seg in seg_list:
        if seg[0]==">":
            inv_seg="<"+seg[1:]
        elif seg[0]=="<":
            inv_seg=">"+seg[1:]
        else:
            inv_seg=seg
        list_inverted.append(inv_seg)
    return list(reversed(list_inverted))