offtracker 2.12.3__zip → 2.13.1__zip

{offtracker-2.12.3/offtracker.egg-info → offtracker-2.13.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: offtracker
-Version: 2.12.3
+Version: 2.13.1
 Summary: Tracking-seq data analysis
 Home-page: https://github.com/Lan-lab/offtracker
 Author: Runda Xu

{offtracker-2.12.3 → offtracker-2.13.1}/offtracker/X_offtracker.py

@@ -2,7 +2,8 @@
 import pandas as pd
 import polars as pl
 import numpy as np
-import os, sys
+import os, sys, re
+import offtracker.X_sequence as xseq
 sys.path.append( os.path.abspath(os.path.dirname(__file__)) )
 
 def fdr(p_vals):
@@ -115,6 +116,29 @@ def target_signal(df_bdg_chr, chrom, cleavage_site, flank_max=100000, smooth_tim
                   binsize=100, flank_regions=[500,1000,2000,5000], 
                   length_bkg = 20000, length_binsize=1000, length_min_noise=0.2, n_std=1, 
                   end='end',start='start',value='residual', pct_offset=0.0):
+    """_summary_
+
+    Args:
+        df_bdg_chr (_type_): .bdg table with the same chromosome
+        chrom (_type_): chr name
+        cleavage_site (_type_): cleavage site
+        flank_max (int, optional): _description_. Defaults to 100000.
+        smooth_times (int, optional): _description_. Defaults to 1.
+        window_size (int, optional): _description_. Defaults to 3.
+        binsize (int, optional): _description_. Defaults to 100.
+        flank_regions (list, optional): _description_. Defaults to [500,1000,2000,5000].
+        length_bkg (int, optional): _description_. Defaults to 20000.
+        length_binsize (int, optional): _description_. Defaults to 1000.
+        length_min_noise (float, optional): _description_. Defaults to 0.2.
+        n_std (int, optional): _description_. Defaults to 1.
+        end (str, optional): _description_. Defaults to 'end'.
+        start (str, optional): _description_. Defaults to 'start'.
+        value (str, optional): _description_. Defaults to 'residual'.
+        pct_offset (float, optional): _description_. Defaults to 0.0.
+
+    Returns:
+        _type_: _description_
+    """
     # 输入数据必须是同一条染色体内的
     # 统计 flank regions 的个数
     # n_regions = len(flank_regions)
@@ -328,6 +352,25 @@ def target_signal(df_bdg_chr, chrom, cleavage_site, flank_max=100000, smooth_tim
 
 def target_signal_chunk(df_bdg_chr, df_alignment_chr, flank_max=100000, smooth_times = 1, window_size = 3, binsize=100, flank_regions=[500,1000,2000,5000], 
                         length_bkg = 20000, length_binsize=1000, length_min_noise=0.2, n_std=1, pct_offset=0.0):
+    """
+
+    Args:
+        df_bdg_chr (_type_): .bdg table with the same chromosome
+        df_alignment_chr (_type_): candidate sites
+        flank_max (int, optional): _description_. Defaults to 100000.
+        smooth_times (int, optional): _description_. Defaults to 1.
+        window_size (int, optional): _description_. Defaults to 3.
+        binsize (int, optional): _description_. Defaults to 100.
+        flank_regions (list, optional): _description_. Defaults to [500,1000,2000,5000].
+        length_bkg (int, optional): _description_. Defaults to 20000.
+        length_binsize (int, optional): _description_. Defaults to 1000.
+        length_min_noise (float, optional): _description_. Defaults to 0.2.
+        n_std (int, optional): _description_. Defaults to 1.
+        pct_offset (float, optional): _description_. Defaults to 0.0.
+
+    Returns:
+        _type_: _description_
+    """
     # 输入数据必须是同一条染色体内的
     list_target_all = []
     for a_row in df_alignment_chr.iterrows():
@@ -348,3 +391,273 @@ def target_signal_chunk(df_bdg_chr, df_alignment_chr, flank_max=100000, smooth_t
     return df_result
 
 
+
+
+
+
+
+
+
+
+
+############################################################################
+# 2025.08.08 新写的的基于 Bio.Align 的 local realign，用于局部修正脱靶位点坐标 #
+############################################################################
+def create_substitution_matrix(mismatch_score=0.01):
+    """
+    Create substitution matrix for DNA alignment using Bio.Align format.
+    """
+    alphabet = 'ATGCN'
+    matrix = np.full((len(alphabet), len(alphabet)), mismatch_score)
+    
+    # Set match scores
+    for i in range(len(alphabet)):
+        matrix[i][i] = 2.0
+    
+    # N matches with everything
+    n_idx = alphabet.index('N')
+    matrix[n_idx, :] = 2.0
+    matrix[:, n_idx] = 2.0
+    
+    return matrix, alphabet
+
+def sgRNA_alignment_new(a_key, sgRNA, seq, substitution_matrix=None, alphabet=None, 
+                   mismatch_score=0.01):
+    """
+    Perform local alignment using Bio.Align instead of deprecated pairwise2.
+    """
+    from Bio import Align
+    if substitution_matrix is None or alphabet is None:
+        substitution_matrix, alphabet = create_substitution_matrix(mismatch_score)
+    
+    # Create aligner
+    aligner = Align.PairwiseAligner()
+    aligner.substitution_matrix = Align.substitution_matrices.Array(
+        alphabet=alphabet, dims=2, data=substitution_matrix
+    )
+    aligner.open_gap_score = -2
+    aligner.extend_gap_score = -2
+    aligner.mode = 'local'
+    
+    try:
+        # Perform alignment
+        alignments = aligner.align(sgRNA, seq)
+
+        if not alignments:
+            # No alignment found, return default values
+            return [0, 0, '', f"{a_key.split(':')[0]}:0-0", 0, 0, len(sgRNA)]
+        
+        # Convert to list for indexing
+        alignments = list(alignments)
+         
+        # Extract alignment information
+        coords = alignments[0].coordinates
+        start_target = coords[1][0]
+        end_target = coords[1][-1]
+        
+        # Extract target sequence directly from coordinates
+        # target = seq[start_target:end_target]
+        
+        # Get aligned sequences for detailed analysis
+        alignment_str = str(alignments[0])
+        alignment_lines = alignment_str.split('\n')
+        if len(alignment_lines) >= 3:
+            aligned_sgrna = [x for x in alignment_lines[0].split(' ') if x != '']
+            aligned_genome = [x for x in alignment_lines[2].split(' ') if x != '']
+        else:
+            raise ValueError("Unexpected alignment format")
+        
+        assert int(aligned_sgrna[-1]) == len(sgRNA)
+        
+        # Calculate indels and mismatches
+        # deletion = RNA bulge
+        # insertion = DNA bulge
+        aligned_sgrna_seq = aligned_sgrna[-2]
+        aligned_genome_seq = aligned_genome[-2]
+        insertion = aligned_sgrna_seq.count('-') if '-' in aligned_sgrna_seq else 0
+        deletion = aligned_genome_seq.count('-') if '-' in aligned_genome_seq else 0
+        
+        # Count mismatches by comparing sequences directly
+        # mismatch = 0
+        # assert len(aligned_sgrna_seq) == len(aligned_genome_seq)
+        # for i in range(len(aligned_sgrna_seq)):
+        #     if (aligned_sgrna_seq[i] != aligned_genome_seq[i]) & (aligned_sgrna_seq[i] != 'N') & (aligned_genome_seq[i] != 'N'):
+        #         mismatch += 1
+
+        mismatch = round((alignments[0].score % 1)/mismatch_score)
+        
+        # Calculate target location
+        pos_st = int(a_key.split('-')[0].split(':')[1]) + 1
+        chr_name = a_key.split(':')[0]
+        target_st = pos_st + start_target
+        target_ed = pos_st + end_target - 1
+        target_location = f"{chr_name}:{target_st}-{target_ed}"
+        
+        score = alignments[0].score
+        
+        return [score, aligned_genome_seq, target_location, deletion, insertion, mismatch]
+            
+    except Exception as e:
+        print(f"Alignment error for {a_key}: {e}")
+        return [0, 0, '', f"{a_key.split(':')[0]}:0-0", 0, 0, len(sgRNA)]
+
+def local_realign(sgRNA_seq, fasta, PAM='NGG', PAM_loc='downstream'):
+    # 添加 PAM
+    if PAM_loc == 'downstream':
+        sgRNA_PAM_fw = sgRNA_seq + PAM
+    else:
+        sgRNA_PAM_fw = PAM + sgRNA_seq
+    sgRNA_PAM_rv = xseq.reverse_complement(sgRNA_PAM_fw)
+    list_args_fw=[]
+    list_args_rv=[]
+    for a_key, a_seq in fasta.items():
+        # 2025.04.25 修正大小写问题
+        a_seq = re.sub('[^ATCG]','N',a_seq.upper())
+        list_args_fw.append( [a_key, sgRNA_PAM_fw, a_seq])
+        list_args_rv.append( [a_key, sgRNA_PAM_rv, a_seq])
+    list_align_forward = [sgRNA_alignment_new(*args) for args in list_args_fw]
+    list_align_reverse = [sgRNA_alignment_new(*args) for args in list_args_rv]
+    #
+    df_align_forward = pd.DataFrame(list_align_forward, columns= ['fw_score', 'fw_target','fw_location','fw_deletion','fw_insertion','fw_mismatch'])
+    df_align_reverse = pd.DataFrame(list_align_reverse, columns= ['rv_score', 'rv_target','rv_location','rv_deletion','rv_insertion','rv_mismatch'])
+    df_align_reverse['rv_target'] = df_align_reverse['rv_target'].apply(xseq.reverse_complement)
+    df_candidate = pd.concat([df_align_forward,df_align_reverse],axis=1)
+    df_candidate['location'] = fasta.keys()
+    df_candidate['alignment_score'] = df_candidate[['fw_score','rv_score']].max(axis=1)
+    df_candidate['best_seq_score'] = df_candidate[['fw_score', 'rv_score']].max(axis=1)
+    df_candidate['best_strand'] = df_candidate[['fw_score', 'rv_score']].idxmax(axis='columns').replace({'fw_score':'+', 'rv_score':'-'})
+    df_candidate.loc[df_candidate['fw_score']==df_candidate['rv_score'],'best_strand']='equal_score'
+            
+    # GG check
+    # 2023.12.05 增加 cleavage_site 推测
+    list_best_target = []
+    list_best_location = []
+    list_cleavage_site = []
+    list_delete = []
+    list_insert = []
+    list_mismat = []
+    list_GG = []
+    for a_row in df_candidate.iterrows():
+        if a_row[1]['best_strand']=='+':
+            list_best_target.append(a_row[1]['fw_target'])
+            list_best_location.append(a_row[1]['fw_location'])
+            list_cleavage_site.append(int(a_row[1]['fw_location'].split('-')[1]) - 6)
+            list_delete.append(a_row[1]['fw_deletion'])
+            list_insert.append(a_row[1]['fw_insertion'])
+            list_mismat.append(a_row[1]['fw_mismatch'])
+            if a_row[1]['fw_target'][-2:]=='GG':
+                list_GG.append('OK')
+            else:
+                list_GG.append('NO')                     
+        elif a_row[1]['best_strand']=='-':
+            list_best_target.append(a_row[1]['rv_target'])
+            list_best_location.append(a_row[1]['rv_location'])
+            list_cleavage_site.append(int(a_row[1]['rv_location'].split('-')[0].split(':')[1]) + 5)
+            list_delete.append(a_row[1]['rv_deletion'])
+            list_insert.append(a_row[1]['rv_insertion'])
+            list_mismat.append(a_row[1]['rv_mismatch'])
+            if a_row[1]['rv_target'][-2:]=='GG':
+                list_GG.append('OK')
+            else:
+                list_GG.append('NO')  
+        else:
+            if a_row[1]['fw_target'][-2:]=='GG':
+                list_best_target.append(a_row[1]['fw_target'])
+                list_best_location.append(a_row[1]['fw_location'])
+                list_cleavage_site.append(int(a_row[1]['fw_location'].split('-')[1]) - 6)
+                list_delete.append(a_row[1]['fw_deletion'])
+                list_insert.append(a_row[1]['fw_insertion'])
+                list_mismat.append(a_row[1]['fw_mismatch'])
+                list_GG.append('OK_same_score')
+            # 发现没有 GG 则看 RC
+            elif a_row[1]['rv_target'][-2:]=='GG':
+                list_best_target.append(a_row[1]['rv_target'])
+                list_best_location.append(a_row[1]['rv_location'])
+                list_cleavage_site.append(int(a_row[1]['rv_location'].split('-')[0].split(':')[1]) + 5)
+                list_delete.append(a_row[1]['rv_deletion'])
+                list_insert.append(a_row[1]['rv_insertion'])
+                list_mismat.append(a_row[1]['rv_mismatch'])
+                list_GG.append('OK_same_score')
+            else:
+                list_best_target.append(a_row[1]['fw_target'])
+                list_best_location.append(a_row[1]['fw_location'])
+                list_cleavage_site.append(int(a_row[1]['fw_location'].split('-')[1]) - 6)
+                list_delete.append(a_row[1]['fw_deletion'])
+                list_insert.append(a_row[1]['fw_insertion'])
+                list_mismat.append(a_row[1]['fw_mismatch'])                    
+                list_GG.append('NO_same_score')
+    # 记入 df_candidate
+    df_candidate['deletion'] = list_delete
+    df_candidate['insertion'] = list_insert
+    df_candidate['mismatch'] = list_mismat
+    df_candidate['GG'] = list_GG
+    df_candidate['best_target'] = list_best_target
+    df_candidate['target_location'] = list_best_location
+    df_candidate['cleavage_site'] = list_cleavage_site
+    df_candidate = pd.concat([xseq.bedfmt(df_candidate['target_location']), df_candidate], axis=1)
+
+    return df_candidate
+
+def left_realign(dp_bdg_chr, loc_shift_left, ref_fasta, sgRNA_seq, PAM, PAM_loc, n_iter):
+    # print(loc_shift_left)
+    fasta = xseq.get_seq(loc_shift_left, ref_fasta)
+    df_candidate = local_realign(sgRNA_seq, fasta, PAM, PAM_loc)
+    sr_candidate = df_candidate.iloc[0].copy()
+    chrom = sr_candidate['chr']
+    cleavage_site = sr_candidate['cleavage_site']
+    flank_regions = [500]
+    signals = target_signal(dp_bdg_chr.to_pandas(), chrom, cleavage_site, flank_regions=flank_regions)
+    L_neg_1000 = signals[2]
+    R_neg_1000 = signals[5]
+    # 如果右侧范围变负数了，说明过头了
+    if R_neg_1000 < 0:
+        sr_candidate.loc['realign'] = 'fail'
+        return sr_candidate
+
+    # 计算左移后的 L_neg_1000，如果还是负数则迭代，最多迭代 10 次
+    if L_neg_1000 < 0:
+        st = sr_candidate['st']
+        ed = sr_candidate['ed']
+        loc_shift_left = f'{chrom}:{int(st)-1000}-{int(ed)-20}'
+        n_iter += 1
+        if n_iter < 10:
+            return left_realign(dp_bdg_chr, loc_shift_left, ref_fasta, sgRNA_seq, PAM, PAM_loc, n_iter)
+        else:
+            sr_candidate.loc['realign'] = 'fail'
+            return sr_candidate
+    else:
+        sr_candidate.loc['realign'] = 'success'
+        return sr_candidate
+
+def right_realign(dp_bdg_chr, loc_shift_right, ref_fasta, sgRNA_seq, PAM, PAM_loc, n_iter):
+    # print(loc_shift_right)
+    fasta = xseq.get_seq(loc_shift_right, ref_fasta)
+    df_candidate = local_realign(sgRNA_seq, fasta, PAM, PAM_loc)
+    sr_candidate = df_candidate.iloc[0].copy()
+    chrom = sr_candidate['chr']
+    cleavage_site = sr_candidate['cleavage_site']
+    flank_regions = [500]
+    signals = target_signal(dp_bdg_chr.to_pandas(), chrom, cleavage_site, flank_regions=flank_regions)
+    L_neg_1000 = signals[2]
+    R_neg_1000 = signals[5]
+    # 如果左侧范围变负数了，说明过头了
+    if L_neg_1000 < 0:
+        sr_candidate.loc['realign'] = 'fail'
+        return sr_candidate
+
+    # 计算右移后的 R_neg_1000，如果还是负数则迭代，最多迭代 10 次
+    if R_neg_1000 < 0:
+        st = sr_candidate['st']
+        ed = sr_candidate['ed']
+        loc_shift_right = f'{chrom}:{int(st)+20}-{int(ed)+1000}'
+        n_iter += 1
+        if n_iter < 10:
+            return right_realign(dp_bdg_chr, loc_shift_right, ref_fasta, sgRNA_seq, PAM, PAM_loc, n_iter)
+        else:
+            sr_candidate.loc['realign'] = 'fail'
+            return sr_candidate
+    else:
+        sr_candidate.loc['realign'] = 'success'
+        return sr_candidate
+
+

{offtracker-2.12.3 → offtracker-2.13.1}/offtracker/X_sequence.py

@@ -3,7 +3,8 @@ import math
 import pandas as pd
 from itertools import product
 import numpy as np
-import os, glob
+import os, glob, sys
+import polars as pl
 
 ambiguous_nt = {'A': ['A'],
                 'T': ['T'],
@@ -246,6 +247,137 @@ def sgRNA_alignment(a_key, sgRNA, seq, frag_len, DNA_matrix=None, mismatch_score
     else:
         return [best_alignment.score, position_pct, target, target_location, deletion, insertion, mismatch]
 
+def sgRNA_alignment_new(a_key, sgRNA, seq, substitution_matrix=None, alphabet=None, 
+                   mismatch_score=0.01):
+    """
+    Perform local alignment using Bio.Align instead of deprecated pairwise2.
+    """
+    if substitution_matrix is None or alphabet is None:
+        substitution_matrix, alphabet = create_substitution_matrix(mismatch_score)
+    
+    # Create aligner
+    aligner = Align.PairwiseAligner()
+    aligner.substitution_matrix = Align.substitution_matrices.Array(
+        alphabet=alphabet, dims=2, data=substitution_matrix
+    )
+    aligner.open_gap_score = -2
+    aligner.extend_gap_score = -2
+    aligner.mode = 'local'
+    
+    try:
+        # Perform alignment
+        alignments = aligner.align(sgRNA, seq)
+
+        if not alignments:
+            # No alignment found, return default values
+            return [0, 0, '', f"{a_key.split(':')[0]}:0-0", 0, 0, len(sgRNA)]
+        
+        # Convert to list for indexing
+        alignments = list(alignments)
+         
+        # Extract alignment information
+        coords = alignments[0].coordinates
+        start_target = coords[1][0]
+        end_target = coords[1][-1]
+        
+        # Extract target sequence directly from coordinates
+        # target = seq[start_target:end_target]
+        
+        # Get aligned sequences for detailed analysis
+        alignment_str = str(alignments[0])
+        alignment_lines = alignment_str.split('\n')
+        if len(alignment_lines) >= 3:
+            aligned_sgrna = [x for x in alignment_lines[0].split(' ') if x != '']
+            aligned_genome = [x for x in alignment_lines[2].split(' ') if x != '']
+        else:
+            raise ValueError("Unexpected alignment format")
+        
+        assert int(aligned_sgrna[-1]) == len(sgRNA)
+        
+        # Calculate indels and mismatches
+        # deletion = RNA bulge
+        # insertion = DNA bulge
+        aligned_sgrna_seq = aligned_sgrna[-2]
+        aligned_genome_seq = aligned_genome[-2]
+        insertion = aligned_sgrna_seq.count('-') if '-' in aligned_sgrna_seq else 0
+        deletion = aligned_genome_seq.count('-') if '-' in aligned_genome_seq else 0
+        
+        # Count mismatches by comparing sequences directly
+        # mismatch = 0
+        # assert len(aligned_sgrna_seq) == len(aligned_genome_seq)
+        # for i in range(len(aligned_sgrna_seq)):
+        #     if (aligned_sgrna_seq[i] != aligned_genome_seq[i]) & (aligned_sgrna_seq[i] != 'N') & (aligned_genome_seq[i] != 'N'):
+        #         mismatch += 1
+
+        mismatch = round((alignments[0].score % 1)/mismatch_score)
+        
+        # Calculate target location
+        pos_st = int(a_key.split('-')[0].split(':')[1]) + 1
+        chr_name = a_key.split(':')[0]
+        target_st = pos_st + start_target
+        target_ed = pos_st + end_target - 1
+        target_location = f"{chr_name}:{target_st}-{target_ed}"
+        
+        score = alignments[0].score
+        
+        return [score, aligned_genome_seq, target_location, deletion, insertion, mismatch]
+            
+    except Exception as e:
+        print(f"Alignment error for {a_key}: {e}")
+        return [0, 0, '', f"{a_key.split(':')[0]}:0-0", 0, 0, len(sgRNA)]
+
+
+def get_seq(location, ref_fasta, return_df=False) -> dict:
+    """
+    根据 genome location 取序列
+    location 如果是 list, str 则形式为 "chr1:123456-123458" 或 ["chr1:123456-123458", "chr2:123456-123458"]
+    location 如果是 pd.DataFrame, pl.DataFrame 则默认前三列为 bed 格式
+    ref_fasta 是参考基因组 fasta 文件的路径
+    默认返回字典, key 是位置, value 是序列
+    如果 return_df 为 True, 则返回 pl.DataFrame, 第一列为位置, 第二列为序列
+
+    pybedtools 返回的序列实际上没有包括坐标 start 的那个碱基，这一点和 twoBitToFa 一样
+    但是 IGV/UCSC 等序列是包括 start 的碱基，blast 的结果也是包括 start 的
+    所以在后续分析时要注意这一点
+    """
+    if sys.platform[:3]=='win':
+        # windows 似乎装不了 pybedtools
+        raise ValueError('windows 似乎装不了 pybedtools')
+    else:
+        import pybedtools
+    #########
+    # 根据 genome location 取序列
+    #########
+    if isinstance(location,(list,str)):
+        bed_loc = bedfmt(location)
+    elif isinstance(location,pd.DataFrame):
+        bed_loc = location.iloc[:,:3]
+    elif isinstance(location,pl.DataFrame):
+        bed_loc = location[:,:3]
+    else:
+        raise ValueError('location must be a list, str or pd.DataFrame')
+    
+    fasta = pybedtools.example_filename(ref_fasta)
+    temp_bed = './temp_amp_loc.bed'
+    write_bed(bed_loc, temp_bed)
+    a = pybedtools.BedTool(temp_bed)
+    a = a.sequence(fi=fasta)
+    with open(a.seqfn, encoding='utf-8') as f:
+        dict_seq = {} # 定义一个空的字典
+        for line in f:
+            line = line.strip() # 去除末尾换行符
+            if line[0] == '>':
+                header = line[1:]
+            else:
+                sequence = line
+                dict_seq[header] = dict_seq.get(header,'') + sequence
+    
+    # remove temp_amp_loc.bed
+    os.remove(temp_bed)
+    if return_df:
+        return pl.DataFrame(list(dict_seq.items()), orient='row', schema={'location':pl.String,'sequence':pl.String})
+    else:
+        return dict_seq
 
 def combine_df(list_df, op = 'mean'):
     # df 行列、结构必须一模一样，非数字部分也一模一样，只有数字不同

{offtracker-2.12.3 → offtracker-2.13.1}/offtracker/_version.py

@@ -1,4 +1,4 @@
-__version__ = "2.12.3"
+__version__ = "2.13.1"
 # 2023.08.11. v1.1.0	adding a option for not normalizing the bw file
 # 2023.10.26. v1.9.0	prerelease for v2.0
 # 2023.10.27. v2.0.0	大更新，还没微调
@@ -33,12 +33,14 @@ __version__ = "2.12.3"
 # 2025.04.25. v2.8.0	修复了 offtracker candidates 会把小写序列转换成 N 的 bug
 # 2025.05.22. v2.9.0	翻新部分代码结构
 # 2025.06.05. v2.10.0	增加了QC模块。保留了负数score的记录，并在plot时显示为红字。增加了 "--ignore_chr" 用于跳过common chr过滤。
-# 2025.06.17. v2.10.7   修复翻新代码结构导致的bug
-# 2025.06.27. v2.10.8   将 chmod 放在了 setup.py 里
-# 2025.06.28. v2.10.9   现在 pip 都是从 wheel 安装，不再运行 setup.py，所以增加一个 offtracker_init.py
-# 2025.06.28. v2.10.10  直接塞 script 里试试
-# 2025.06.28. v2.10.11  回滚到2.10.9外加修正
-# 2025.07.02. v2.11.4   基于 blast 的缺陷更新 candidates，去除 quick mode
-# 2025.07.04. v2.11.5   offtracker_analysis 提前 skip 已有结果的样本
-# 2025.07.04. v2.12.2   新增 region_index 标记区域，用于更好的去重
-# 2025.07.18. v2.12.3   新增QC自动避免重复读取 trimmed fastq files
+# 2025.06.17. v2.10.7	修复翻新代码结构导致的bug
+# 2025.06.27. v2.10.8	将 chmod 放在了 setup.py 里
+# 2025.06.28. v2.10.9	现在 pip 都是从 wheel 安装，不再运行 setup.py，所以增加一个 offtracker_init.py
+# 2025.06.28. v2.10.10	直接塞 script 里试试
+# 2025.06.28. v2.10.11	回滚到2.10.9外加修正
+# 2025.07.02. v2.11.4	基于 blast 的缺陷更新 candidates，去除 quick mode
+# 2025.07.04. v2.11.5	offtracker_analysis 提前 skip 已有结果的样本
+# 2025.07.04. v2.12.2	新增 region_index 标记区域，用于更好的去重
+# 2025.07.18. v2.12.3	新增QC自动避免重复读取 trimmed fastq files
+# 2025.08.08. v2.13.0	测试 local realign 功能
+# 2025.08.09. v2.13.1	测试 correction 功能

{offtracker-2.12.3 → offtracker-2.13.1/offtracker.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: offtracker
-Version: 2.12.3
+Version: 2.13.1
 Summary: Tracking-seq data analysis
 Home-page: https://github.com/Lan-lab/offtracker
 Author: Runda Xu

{offtracker-2.12.3 → offtracker-2.13.1}/offtracker.egg-info/SOURCES.txt

@@ -24,6 +24,7 @@ offtracker/utility/offtracker_blacklist_mm10.merged.bed
 scripts/offtracker_analysis.py
 scripts/offtracker_candidates.py
 scripts/offtracker_config.py
+scripts/offtracker_correction.py
 scripts/offtracker_init.py
 scripts/offtracker_plot.py
 scripts/offtracker_qc.py

{offtracker-2.12.3 → offtracker-2.13.1}/scripts/offtracker_analysis.py

@@ -22,24 +22,29 @@ def main():
     parser = argparse.ArgumentParser()
     parser.description='Analyze the Tracking-seq data.'
     parser.add_argument('-f','--folder'  , type=str, required=True,    nargs='+', help='Directory of the data folder.' )
-    parser.add_argument('--seqfolder'    , type=str, required=True,    help='folder containing df_candidate created by offtracker_cadidates.py.')
+    parser.add_argument('--seqfolder'    , type=str, default ='none',  help='folder containing df_candidate created by offtracker_cadidates.py.')
     parser.add_argument('--name'         , type=str, required=True,    help='custom name of the sgRNA' )
     parser.add_argument('--exp'          , type=str, default='all',    nargs='+', help='A substring mark in the name of experimental samples. The default is to use all samples other than control' )
     parser.add_argument('--control'      , type=str, default='none',   nargs='+', help='A substring mark in the name of control samples. The default is no control. "others" for all samples other than --exp.' )
-    parser.add_argument('--fdr'          , type=float, default=0.05,     help='FDR threshold for the final result. Default is 0.05.')
-    parser.add_argument('--score'        , type=float, default=1.9,        help='Track score threshold for the final result. Default is 1.9.')
+    parser.add_argument('--fdr'          , type=float, default=0.05,   help='FDR threshold for the final result. Default is 0.05.')
+    parser.add_argument('--score'        , type=float, default=1.9,    help='Track score threshold for the final result. Default is 1.9.')
     parser.add_argument('--smooth'       , type=int, default=1,        help='Smooth strength for the signal.')
     parser.add_argument('--window'       , type=int, default=3,        help='Window size for smoothing the signal.')
     parser.add_argument('--binsize'      , type=int, default=100,      help='Window size for smoothing the signal.')
     parser.add_argument('--flank_max'    , type=int, default=100000,   help='Maximun flanking distance from the candidate site.')
     parser.add_argument('--flank_regions', type=int, default=[1000,2000,3000,5000], nargs='+',help='flanking regions for calculating signal.')
     parser.add_argument('--SeqScorePower', type=float, default=4,      help='The seq score power' )
-    parser.add_argument('--CtrClip'      , type=float, default=-0.5,     help='The lower clip for control samples' )
+    parser.add_argument('--CtrClip'      , type=float, default=-0.5,   help='The lower clip for control samples' )
     parser.add_argument('-t','--thread'  , type=int, default=4,        help='Number of threads for parallel computing')
     parser.add_argument('-g','--genome'  , type=str, default='hg38',   help='File of chromosome sizes, or "hg38", "mm10" ')
     parser.add_argument('-o','--outdir'  , type=str, default='first',  help='The output folder. Default is the first folder of --folder' )
     parser.add_argument('--outname'      , type=str, default='same',   help='The suffix of output files. Default is the same --exp' )
     parser.add_argument('--signal_only'  , action='store_true', help='A developer option: stop before group analysis. ' )
+    # for offtracker_correction
+    parser.add_argument('--check_loc'    , action='store_true', help='New in v2.13, for other scripts. Do not use this option. ' )
+    parser.add_argument('--seqfile'      , type=str, default='none',   help='Assign a specific df_candidate file.')
+
+    # other parameters
     # parser.add_argument('--individual_results', action='store_true', help='When multiple samples meet the exp pattern, only one merged result is generated.\n' \
     #                                                                       'Set --individual_results to additionally output the individual result of each exp sample. ' )
     parser.add_argument('--overwrite'    , action='store_true', help='Whether to overwrite existed dataframes.' )
@@ -79,11 +84,16 @@ def main():
     
     # load df_candidate
     try:
-        df_candidate = pl.read_csv(os.path.join(args.seqfolder,f'df_candidate_{sgRNA_name}.csv')).to_pandas()
+        if args.seqfile != 'none':
+            df_candidate = pl.read_csv(args.seqfile).to_pandas()
+        elif args.seqfolder != 'none':
+            df_candidate = pl.read_csv(os.path.join(args.seqfolder,f'df_candidate_{sgRNA_name}.csv')).to_pandas()
+        else:
+            raise ValueError('Please provide --seqfolder or --seqfile')
         df_candidate.index = df_candidate['target_location']
         df_candidate_brief = df_candidate[['chr','st','ed','best_strand','best_target','best_seq_score',
                                  'deletion', 'insertion','mismatch', 'GG', 
-                                 'target_location', 'cleavage_site', 'ID_1','ID_2', 'region_index']] # 2025.07.06 添加 region_index
+                                 'target_location', 'cleavage_site', 'region_index']] # 2025.07.06 添加 region_index, 去除 'ID_1','ID_2',
         df_candidate_sub = df_candidate[['chr','cleavage_site']]
     except FileNotFoundError:
         return 'Please run offtracker_candidates.py first and provide the correct directory with --seqfolder'
@@ -160,8 +170,11 @@ def main():
         if (os.path.isfile(output))&(not args.overwrite):
             print(output, 'exists, skipped')
             continue
-        df_bdg = xseq.read_bed(a_file)
-        df_bdg.columns = ['chr','start','end','residual']
+        # 2025.08.09. 改用 pl 读取加速
+        df_bdg = pl.read_csv(a_file, separator='\t', has_header=False,
+                             schema_overrides={'chr':pl.String,'start':pl.Int32,
+                                               'end':pl.Int32,'residual':pl.Float32}).to_pandas() # xseq.read_bed(a_file)
+        # df_bdg.columns = ['chr','start','end','residual']
         # 将 df_bdg 按照染色体分组
         sample_groups = df_bdg.groupby('chr')
         # 2024.06.03. fix a bug that df_bdg has less chr than df_candidate
@@ -308,7 +321,7 @@ def main():
         # 2025.07.06 更新去重方式
         df_result = df_score.drop_duplicates(subset=['region_index'], keep='first').copy()
 
-        # 标准化分布      
+        # 标准化分布，2025.08.09
         target_std=0.15
         n_outliers = int(np.ceil(len(df_result)*0.01))
         score_bkg = df_result['raw_score'][n_outliers:-n_outliers]
@@ -317,7 +330,7 @@ def main():
         df_result['track_score'] = (df_result['raw_score'] - mean_score_bkg) / std_score_bkg
         df_result['track_score'] = df_result['track_score']*target_std + 1
         df_result = df_result.sort_values(by='track_score', ascending=False)
-        df_result['log2_track_score'] = np.log2(df_result['track_score'].clip(lower=0.5))   
+        df_result['log2_track_score'] = np.log2(df_result['track_score'].clip(lower=0.5))
 
         # 单边信号周围有更高分的，去掉
         # v2.1 后 cols_L, cols_R 要手动
@@ -362,28 +375,30 @@ def main():
         df_result['rank'] = range(1,len(df_result)+1)
         df_result.to_csv(output)
 
-    output = f'Offtracker_result_{outname}.csv'
-    # 2024.06.03. 以防 fdr<=fdr_thresh 滤掉了 track_score>=2 的位点
-    bool_fdr = df_result['fdr']<=fdr_thresh
-    bool_score = df_result['track_score']>=score_thresh
-    # 2025.06.05. BE可能会形成单边信号，导致 track_score 为负数，也保留
-    bool_neg_score = df_result['track_score']< -1
-    df_output = df_result[bool_fdr|bool_score|bool_neg_score].copy()
-    if pattern_ctr != 'none':
-        df_output = df_output[['target_location', 'best_strand','best_target','deletion','insertion','mismatch',
-                            'exp_L_length', 'exp_R_length','ctr_L_length','ctr_R_length','L_length','R_length','signal_length',
-                            'norm_best_seq_score','track_score', 'log2_track_score','fdr','rank']]
-        df_output.columns = ['target_location', 'strand', 'target', 'deletion', 'insertion', 'mismatch', 
-                            'exp_L_length', 'exp_R_length','ctr_L_length','ctr_R_length','L_length','R_length','signal_length',
-                            'seq_score', 'track_score', 'log2_track_score','FDR', 'rank']
-    else:
-        df_output = df_output[['target_location', 'best_strand','best_target','deletion','insertion','mismatch',
-                            'L_length', 'R_length','signal_length',
-                            'norm_best_seq_score','track_score', 'log2_track_score','fdr','rank']]
-        df_output.columns = ['target_location', 'strand', 'target', 'deletion', 'insertion', 'mismatch', 
-                            'L_length', 'R_length','signal_length',
-                            'seq_score', 'track_score', 'log2_track_score','FDR', 'rank']
-    df_output.to_csv(f'Offtracker_result_{outname}.csv', index=False)
+    if not args.check_loc:
+        output = f'Offtracker_result_{outname}.csv'
+        # 2024.06.03. 以防 fdr<=fdr_thresh 滤掉了 track_score>=2 的位点
+        bool_fdr = df_result['fdr']<=fdr_thresh
+        bool_score = df_result['track_score']>=score_thresh
+        # 2025.06.05. BE可能会形成单边信号，但是很少见，如果 control 用的是别的 sgRNA 的样本，对应脱靶位置附近一般就是负数
+        # bool_neg_score = df_result['track_score']< -1
+        df_output = df_result[bool_fdr|bool_score].copy()
+        if pattern_ctr != 'none':
+            df_output = df_output[['target_location', 'best_strand','best_target','deletion','insertion','mismatch',
+                                'exp_L_length', 'exp_R_length','ctr_L_length','ctr_R_length','L_length','R_length','signal_length',
+                                'norm_best_seq_score','track_score', 'log2_track_score','fdr','rank']]
+            df_output.columns = ['target_location', 'strand', 'target', 'deletion', 'insertion', 'mismatch',
+                                'exp_L_length', 'exp_R_length','ctr_L_length','ctr_R_length','L_length','R_length','signal_length',
+                                'seq_score', 'track_score', 'log2_track_score','FDR', 'rank']
+        else:
+            df_output = df_output[['target_location', 'best_strand','best_target','deletion','insertion','mismatch',
+                                'L_length', 'R_length','signal_length',
+                                'norm_best_seq_score','track_score', 'log2_track_score','fdr','rank']]
+            df_output.columns = ['target_location', 'strand', 'target', 'deletion', 'insertion', 'mismatch',
+                                'L_length', 'R_length','signal_length',
+                                'seq_score', 'track_score', 'log2_track_score','FDR', 'rank']
+    
+        df_output.to_csv(f'Offtracker_result_{outname}.csv', index=False)
 
     if args.clean:
         shutil.rmtree('./temp')

offtracker-2.13.1/scripts/offtracker_correction.py

@@ -0,0 +1,282 @@
+
+
+import polars as pl
+import pandas as pd
+import numpy as np
+import offtracker
+import argparse
+import os, glob
+import shlex, subprocess
+from scipy.stats import norm
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.description='New function in 2026. Check and correct potential incorrect target locations.'
+    parser.add_argument('-f','--folder'  , type=str, required=True,    nargs='+', help='Directory of the data folder.' )
+    parser.add_argument('--name'         , type=str, required=True,    help='custom name of the sgRNA' )
+    parser.add_argument('--exp'          , type=str, default='all',    nargs='+', help='A substring mark in the name of experimental samples. The default is to use all samples other than control' )
+    parser.add_argument('--control'      , type=str, default='none',   nargs='+', help='A substring mark in the name of control samples. The default is no control. "others" for all samples other than --exp.' )
+    parser.add_argument('--fdr'          , type=float, default=0.05,   help='FDR threshold for the final result. Default is 0.05.')
+    parser.add_argument('--score'        , type=float, default=1.9,    help='Track score threshold for the final result. Default is 1.9.')
+    parser.add_argument('--smooth'       , type=int, default=1,        help='Smooth strength for the signal.')
+    parser.add_argument('--window'       , type=int, default=3,        help='Window size for smoothing the signal.')
+    parser.add_argument('--binsize'      , type=int, default=100,      help='Window size for smoothing the signal.')
+    parser.add_argument('--flank_max'    , type=int, default=100000,   help='Maximun flanking distance from the candidate site.')
+    parser.add_argument('--flank_regions', type=int, default=[1000,2000,3000,5000], nargs='+',help='flanking regions for calculating signal.')
+    parser.add_argument('--SeqScorePower', type=float, default=4,      help='The seq score power' )
+    parser.add_argument('--CtrClip'      , type=float, default=-0.5,   help='The lower clip for control samples' )
+    parser.add_argument('-t','--thread'  , type=int, default=4,        help='Number of threads for parallel computing')
+    parser.add_argument('-g','--genome'  , type=str, default='hg38',   help='File of chromosome sizes, or "hg38", "mm10" ')
+    parser.add_argument('-o','--outdir'  , type=str, default='first',  help='The output folder. Default is the first folder of --folder' )
+    parser.add_argument('--outname'      , type=str, default='same',   help='The suffix of output files. Default is the same --exp' )
+    # new argument
+    parser.add_argument('-r','--ref'     , type=str, required=True,    help='The fasta file of reference genome')
+    parser.add_argument('--sgrna' ,        type=str, required=True,    help='One sgRNA sequence without PAM' )
+    parser.add_argument('--pam'   ,        type=str, required=True,    help='The protospacer adjacent motif' )
+    parser.add_argument('--pam_location',  type=str, default='downstream', help='Upstream or downstream, default is downstream (Cas9)' )
+    # not used
+    parser.add_argument('--seqfolder'    , type=str, required=True,    help='Actually not used in this script.Only in case you forget to remove this argument.')
+
+    args = parser.parse_args()
+    # 2025.08.08. 增加对阳性位点的 target_location 重比对功能，避免 blast 比对后的 realign 在更大范围内的存在不准确的情况
+    # 实验性功能，如果 exp 有多个样本的话目前只取第一个 bdg 来分析
+
+    ##########################
+    ## parameter initiation ##
+    ##########################
+
+    folders = args.folder
+    sgRNA_name = args.name + '_loc_correction'
+    pattern_exp = args.exp
+    pattern_ctr = args.control
+    fdr_thresh = args.fdr
+    score_thresh = args.score
+    binsize = args.binsize
+    flank_max = args.flank_max
+    flank_regions = args.flank_regions
+    smooth_times = args.smooth
+    window_size = args.window
+    seq_score_power = args.SeqScorePower
+    ctr_clip = args.CtrClip
+
+
+    if args.outname == 'same':
+        if isinstance(pattern_exp, list):
+            outname = '_'.join(pattern_exp)
+        else:
+            outname = pattern_exp
+    else:
+        outname = args.outname
+
+    outdir = args.outdir
+    if outdir == 'first':
+        outdir = folders[0]
+    os.chdir(outdir)
+    # out temp folder
+    if not os.path.exists( os.path.join(outdir,'temp') ):
+        os.makedirs(os.path.join(outdir,'temp'))
+    # data temp folder
+    for a_folder in folders:
+        temp_dir = os.path.join(a_folder, 'temp')
+        if not os.path.exists( temp_dir ):
+            os.makedirs(temp_dir)
+
+    ##################
+    ## glob samples ##
+    ##################
+    all_sample_names = []
+    all_sample_files = []
+    for a_folder in folders:    
+        bdg_files = pd.Series(glob.glob(os.path.join( a_folder, '*.add.bdg' ))).sort_values().reset_index(drop=True)
+        sample_names = bdg_files.apply(os.path.basename).str.extract(r'(.*)\.\d+\.add\.bdg',expand=False)
+        all_sample_names.extend( sample_names )
+        all_sample_files.extend( bdg_files )
+    all_sample_files = pd.Series(all_sample_files)
+    all_sample_names = pd.Series(all_sample_names)
+    print('all sample names in the folders:')
+    print(all_sample_names)
+    print('your string pattern for experimental groups: ', pattern_exp)
+    ctr_samples = []
+    if pattern_ctr == 'none':
+        if pattern_exp == 'all':
+            exp_samples = list( all_sample_names )
+        else:
+            exp_samples = []
+            for a_mark in pattern_exp:
+                exp_samples.extend( list( all_sample_names[all_sample_names.str.contains(a_mark)] ) )
+    elif pattern_ctr == 'others':
+        if pattern_exp == 'all':
+            exp_samples = list( all_sample_names )
+        else:
+            exp_samples = []
+            for a_mark in pattern_exp:
+                exp_samples.extend( list( all_sample_names[all_sample_names.str.contains(a_mark)] ) )
+            ctr_samples = list( all_sample_names[~all_sample_names.isin(exp_samples)] )
+    else:
+        for a_mark in pattern_ctr:
+            ctr_samples.extend( list( all_sample_names[all_sample_names.str.contains(a_mark)] ) )
+        if pattern_exp == 'all':
+            exp_samples = list( all_sample_names[~all_sample_names.isin(ctr_samples)] )
+        else:
+            exp_samples = []
+            for a_mark in pattern_exp:
+                exp_samples.extend( list( all_sample_names[all_sample_names.str.contains(a_mark)] ) )
+    n_exp = len(exp_samples)
+    n_ctr = len(ctr_samples)
+    print(f'Experimental group has {n_exp} samples:\n{exp_samples}')
+    print(f'Control group has {n_ctr} samples:\n{ctr_samples}')
+
+    # mark 错误时
+    assert n_exp > 0, 'No experimental sample is found. Please check the name pattern.'
+    if (n_ctr==0)&(pattern_ctr != 'none'):
+        print('Name pattern for control sample(s) was given, but no file meet the pattern.')
+        return 'Program terminated'
+
+    # summarize samples
+    bool_exp = all_sample_names.isin(exp_samples)
+    bool_ctr = all_sample_names.isin(ctr_samples)
+    exp_sample_files = all_sample_files[bool_exp]
+    ctr_sample_files = all_sample_files[bool_ctr]
+    exp_sample_names = all_sample_names[bool_exp]
+    ctr_sample_names = all_sample_names[bool_ctr]
+    selected_sample_files = pd.concat([exp_sample_files,ctr_sample_files])
+    selected_sample_names = pd.concat([exp_sample_names,ctr_sample_names]) # no use
+
+
+
+    ####################
+    ## run correction ##
+    ####################
+
+    # new parameters
+    ref_fasta = args.ref
+    sgRNA_seq = args.sgrna
+    PAM = args.pam
+    PAM_loc = args.pam_location
+    # read result
+    dp_result = pl.read_csv(f'Offtracker_result_{outname}.csv')
+    dp_bdg = pl.read_parquet(selected_sample_files[0], separator='\t', has_header=False,
+                             schema_overrides={'chr':pl.String,'start':pl.Int32,'end':pl.Int32,'residual':pl.Float32})
+    # check and realign
+    bool_left_neg=(dp_result['exp_L_neg_1000']<-5)&(dp_result['exp_R_neg_1000']==0)
+    bool_right_neg=(dp_result['exp_R_neg_1000']<-5)&(dp_result['exp_L_neg_1000']==0)
+    list_good_result = []
+    list_bad_left = []
+    list_bad_right = []
+    for a_left_bool, a_right_bool, a_row in zip(bool_left_neg, bool_right_neg, dp_result.iter_rows(named=True)):
+        if a_left_bool & a_right_bool:
+            raise ValueError('abnormal on both left and right')
+        if a_left_bool:
+            print('left')
+            loc_shift_left = a_row['chr'] + ':' + str(a_row['st']-1000) + '-' + str(a_row['ed']-20)
+            region_index = a_row['region_index']
+            dp_bdg_chr = dp_bdg.filter(pl.col('chr') == a_row['chr'])
+            sr_candidate = offtracker.left_realign(dp_bdg_chr, loc_shift_left, ref_fasta, sgRNA_seq, PAM, PAM_loc, n_iter=0)
+            sr_candidate.loc['region_index'] = region_index
+            list_bad_left.append(sr_candidate)
+        elif a_right_bool:
+            print('right')
+            loc_shift_right = a_row['chr'] + ':' + str(a_row['st']+20) + '-' + str(a_row['ed']+1000)
+            region_index = a_row['region_index']
+            dp_bdg_chr = dp_bdg.filter(pl.col('chr') == a_row['chr'])
+            sr_candidate = offtracker.right_realign(dp_bdg_chr, loc_shift_right, ref_fasta, sgRNA_seq, PAM, PAM_loc, n_iter=0)
+            sr_candidate.loc['region_index'] = region_index
+            list_bad_right.append(sr_candidate)
+        else:
+            list_good_result.append(a_row)
+    dp_result_good = pl.DataFrame(list_good_result)
+    df_cand_left = pd.DataFrame(list_bad_left)
+    df_cand_right = pd.DataFrame(list_bad_right)
+    df_cand_realign = pd.concat([df_cand_left, df_cand_right])
+
+    seqfile = rf'correction_df_candidate_{outname}_realign.csv'
+    df_cand_realign.to_csv(seqfile)
+    
+    # run offtracker_analysis with check_loc mode
+    running_log = rf'correction_analysis_{outname}.log'
+    with open(running_log, "w+") as running_log:
+        command   = f'offtracker_analysis.py -t {args.thread} -g {args.genome} --seqfile {seqfile} --name {sgRNA_name} \
+        --exp {pattern_exp} --control {pattern_ctr} --outname {outname}_loc_correction -f {folders} -o {outdir} \
+        --fdr {fdr_thresh} --window {window_size} --smooth {smooth_times} --SeqScorePower {seq_score_power} \
+        --score {score_thresh} --binsize {binsize} --flank_max {flank_max} --flank_regions {flank_regions} --CtrClip {ctr_clip} \
+        --check_loc'
+        command2  = shlex.split('bash -c "{}"'.format(command))
+        process_1 = subprocess.Popen(command2, stdout=running_log, stderr=subprocess.STDOUT )
+        process_1.wait(timeout=100000)
+        retc  = process_1.returncode
+        if retc==0:
+            print((f'correction_analysis {outname} is done!'))
+        else:
+            print((f'correction_analysis {outname} is failed!'))
+
+
+    #######################
+    ## recalculate score ##
+    #######################
+    dp_result_bkg = pl.read_csv(f'./temp/df_result_{outname}.csv')
+    bool_fdr_bkg = dp_result_bkg['fdr']>fdr_thresh
+    bool_score_bkg = dp_result_bkg['track_score']<score_thresh
+    dp_result_bkg = dp_result_bkg.filter(bool_fdr_bkg & bool_score_bkg)
+    dp_result_realign = pl.read_csv(f'./temp/df_result_{outname}_loc_correction.csv')
+
+    # 兼容旧版输出列名
+    list_col = dp_result_realign.columns[:-5]
+    dp_result_new = pl.concat([dp_result_realign[list_col], dp_result_good[list_col], dp_result_bkg[list_col]])
+
+    # 标准化分布, polars 版   
+    target_std=0.15
+    n_outliers = int(np.ceil(len(dp_result_new)*0.01))
+    score_bkg = dp_result_new['raw_score'][n_outliers:-n_outliers]
+    mean_score_bkg = score_bkg.mean()
+    std_score_bkg = score_bkg.std()
+    dp_result_new = dp_result_new.with_columns(
+        (pl.col('raw_score').sub(mean_score_bkg)/std_score_bkg).alias('track_score')
+    )
+    dp_result_new = dp_result_new.with_columns(
+        pl.col('track_score').mul(target_std).add(1).alias('track_score')
+    )
+    dp_result_new = dp_result_new.with_columns(
+        pl.col('track_score').clip(lower_bound=0.5).log(base=2).alias('log2_track_score')
+    )
+    dp_result_new = dp_result_new.sort('track_score', descending=True)
+
+    # pv and fdr
+    score_for_fitting = dp_result_new['log2_track_score'][n_outliers:-n_outliers]
+    mu, std = norm.fit(score_for_fitting) 
+    print('mean_score:{:.3f};std:{:.3f}'.format(mu,std))
+    dp_result_new = dp_result_new.with_columns(
+        pl.col('log2_track_score').map_elements( lambda x: norm.sf(x,loc=mu,scale=std), return_dtype=pl.Float64 ).clip(lower_bound=1e-320).alias('pv')
+    )
+    dp_result_new = dp_result_new.with_columns(
+        fdr=offtracker.fdr(dp_result_new['pv']).alias('fdr'),
+        rank=pl.Series(range(1,len(dp_result_new)+1))
+    ) #.with_row_index(name='rank',offset=1)
+    dp_result_new.write_csv(f'./temp/df_result_{outname}.csv') # 覆盖原结果
+
+    # ouput Offtracker result
+    bool_fdr = pl.col('fdr')<=fdr_thresh
+    bool_score = pl.col('track_score')>=score_thresh
+    dp_output = dp_result_new.filter(bool_fdr|bool_score).copy()
+    if pattern_ctr != 'none':
+        dp_output = dp_output[['target_location', 'best_strand','best_target','deletion','insertion','mismatch',
+                            'exp_L_length', 'exp_R_length','ctr_L_length','ctr_R_length','L_length','R_length','signal_length',
+                            'norm_best_seq_score','track_score', 'log2_track_score','fdr','rank']]
+        dp_output.columns = ['target_location', 'strand', 'target', 'deletion', 'insertion', 'mismatch',
+                            'exp_L_length', 'exp_R_length','ctr_L_length','ctr_R_length','L_length','R_length','signal_length',
+                            'seq_score', 'track_score', 'log2_track_score','FDR', 'rank']
+    else:
+        dp_output = dp_output[['target_location', 'best_strand','best_target','deletion','insertion','mismatch',
+                            'L_length', 'R_length','signal_length',
+                            'norm_best_seq_score','track_score', 'log2_track_score','fdr','rank']]
+        dp_output.columns = ['target_location', 'strand', 'target', 'deletion', 'insertion', 'mismatch',
+                            'L_length', 'R_length','signal_length',
+                            'seq_score', 'track_score', 'log2_track_score','FDR', 'rank']
+        dp_output.write_csv(f'Offtracker_result_{outname}.csv')
+
+    return 'correction finished'
+
+if __name__ == '__main__' :
+    result = main()
+    print(result)
+
+

{offtracker-2.12.3 → offtracker-2.13.1}/setup.py

@@ -54,6 +54,7 @@ setup(
                 'scripts/offtracker_config.py',
                 'scripts/offtracker_candidates.py',
                 'scripts/offtracker_analysis.py',
+                'scripts/offtracker_correction.py',
                 'scripts/offtracker_plot.py'],
     install_requires=REQUIRED,
     include_package_data=True