mgnify-pipelines-toolkit 0.1.4__py3-none-any.whl → 0.1.6__py3-none-any.whl

mgnify_pipelines_toolkit/analysis/amplicon/assess_inflection_point_mcp.py

@@ -20,28 +20,43 @@ from Bio.Seq import Seq
 import numpy as np
 import pandas as pd
 
-from mgnify_pipelines_toolkit.analysis.amplicon.amplicon_utils import get_read_count, build_cons_seq, build_mcp_cons_dict_list, fetch_mcp
+from mgnify_pipelines_toolkit.analysis.amplicon.amplicon_utils import (
+    get_read_count,
+    build_cons_seq,
+    build_mcp_cons_dict_list,
+    fetch_mcp,
+)
 from mgnify_pipelines_toolkit.constants.thresholds import MCP_MAX_LINE_COUNT
 
+
 def parse_args():
 
     parser = argparse.ArgumentParser()
 
-    parser.add_argument("-i", "--input", required=True, type=str, help="Path to fastq file to choose inflection point")
-    parser.add_argument("-p", "--points", required=True, type=str, help="Path to inflection points file")
+    parser.add_argument(
+        "-i",
+        "--input",
+        required=True,
+        type=str,
+        help="Path to fastq file to choose inflection point",
+    )
+    parser.add_argument(
+        "-p", "--points", required=True, type=str, help="Path to inflection points file"
+    )
     parser.add_argument("-s", "--sample", required=True, type=str, help="Sample ID")
     parser.add_argument("-o", "--output", required=True, type=str, help="Output path")
 
     args = parser.parse_args()
-    
-    _PATH = args.input
-    _POINTS = args.points
-    _SAMPLE = args.sample
-    _OUTPUT = args.output
 
-    return _PATH, _POINTS, _SAMPLE, _OUTPUT
+    path = args.input
+    points = args.points
+    sample = args.sample
+    output = args.output
+
+    return path, points, sample, output
 
-def assess_inflection_point_mcp_for_sample(_PATH, inf_point_list, rev=False):
+
+def assess_inflection_point_mcp_for_sample(path, inf_point_list, rev=False):
     """
     Assess inflection point list, selecting one for automatic primer trimming.
 
@@ -54,17 +69,19 @@ def assess_inflection_point_mcp_for_sample(_PATH, inf_point_list, rev=False):
 
     Returns the cutoff point and the consensus sequence 'forming' the automatically predicted primer
     """
-    
+
     # TODO error handle for empty inflection point list
 
-    start_confs = [] # pre-inf point conservations
-    end_confs = [] # post-inf point conservations
-    start_cons_lens = [] # list for storing lengths of pre-inflection point sequences
-    cons_seq_list = [] # list for storing consensus sequences pre-inflection points
+    start_confs = []  # pre-inf point conservations
+    end_confs = []  # post-inf point conservations
+    start_cons_lens = []  # list for storing lengths of pre-inflection point sequences
+    cons_seq_list = []  # list for storing consensus sequences pre-inflection points
 
-    do_not_include_list = [ i + 5 for i in inf_point_list ] # ignore conservation of inflection point in calculation
+    do_not_include_list = [
+        i + 5 for i in inf_point_list
+    ]  # ignore conservation of inflection point in calculation
 
-    read_count = get_read_count(_PATH) # get readcount from fastq
+    read_count = get_read_count(path)  # get readcount from fastq
 
     max_line_count = None
     if read_count > MCP_MAX_LINE_COUNT:
@@ -72,89 +89,126 @@ def assess_inflection_point_mcp_for_sample(_PATH, inf_point_list, rev=False):
 
     n_prop = 0.8
 
-    for start in inf_point_list: # Looping through the pre-inflection point mcps
-        mcp_len = start + 4 # length of pre-inf mcps is inflection point + 4
-
-        mcp_count_dict = fetch_mcp(_PATH, mcp_len, rev=rev, max_line_count=max_line_count) # get MCP count dict 
-        mcp_cons_list = build_mcp_cons_dict_list(mcp_count_dict, mcp_len) # list of base conservation dicts for mcps
-        cons_seq, cons_confs = build_cons_seq(mcp_cons_list, read_count, n_prop, do_not_include_list, max_line_count=max_line_count) # get list of max base conservations for each index
-                                                                                                    # also get consensus sequence
+    for start in inf_point_list:  # Looping through the pre-inflection point mcps
+        mcp_len = start + 4  # length of pre-inf mcps is inflection point + 4
+
+        mcp_count_dict = fetch_mcp(
+            path, mcp_len, rev=rev, max_line_count=max_line_count
+        )  # get MCP count dict
+        mcp_cons_list = build_mcp_cons_dict_list(
+            mcp_count_dict, mcp_len
+        )  # list of base conservation dicts for mcps
+        cons_seq, cons_confs = build_cons_seq(
+            mcp_cons_list,
+            read_count,
+            n_prop,
+            do_not_include_list,
+            max_line_count=max_line_count,
+        )  # get list of max base conservations for each index
+        # also get consensus sequence
         cons_seq_list.append(cons_seq)
         start_confs.append(np.mean(cons_confs))
         start_cons_lens.append(len(cons_seq))
 
-    for i, end in enumerate(inf_point_list): # Looping through the post-inflection point mcps
-        mcp_len = end + 5 # length of pre-inf mcps is inflection point + 5
-        subs_len = start_cons_lens[i] # length of respective pre-inf point sequence
-        l = mcp_len + subs_len - 1 # final index of MCP
+    for i, end in enumerate(
+        inf_point_list
+    ):  # Looping through the post-inflection point mcps
+        mcp_len = end + 5  # length of pre-inf mcps is inflection point + 5
+        subs_len = start_cons_lens[i]  # length of respective pre-inf point sequence
+        l = mcp_len + subs_len - 1  # final index of MCP
 
-        mcp_count_dict = fetch_mcp(_PATH, l, mcp_len, rev=rev, max_line_count=max_line_count)
+        mcp_count_dict = fetch_mcp(
+            path, l, mcp_len, rev=rev, max_line_count=max_line_count
+        )
         mcp_cons_list = build_mcp_cons_dict_list(mcp_count_dict, subs_len)
-        cons_seq, cons_confs = build_cons_seq(mcp_cons_list, read_count, n_prop, do_not_include_list, subs_len, max_line_count=max_line_count)
+        cons_seq, cons_confs = build_cons_seq(
+            mcp_cons_list,
+            read_count,
+            n_prop,
+            do_not_include_list,
+            subs_len,
+            max_line_count=max_line_count,
+        )
 
         end_confs.append(np.mean(cons_confs))
 
-    diff_res = [ start_confs[i] - end_confs[i] for i in range(len(start_confs))] # get differences between pre- and -post avg conservation values
-    diff_res_sorted = sorted(diff_res, reverse=True) # sort differences from highest to lowest
+    diff_res = [
+        start_confs[i] - end_confs[i] for i in range(len(start_confs))
+    ]  # get differences between pre- and -post avg conservation values
+    diff_res_sorted = sorted(
+        diff_res, reverse=True
+    )  # sort differences from highest to lowest
 
-    ini_max_res = diff_res_sorted[0] # maximum differences
-    curr_max_index = diff_res.index(ini_max_res) # index of maximum differences
+    ini_max_res = diff_res_sorted[0]  # maximum differences
+    curr_max_index = diff_res.index(ini_max_res)  # index of maximum differences
 
-    for res in diff_res_sorted[1:]: # Loop through the rest of the differences
-        curr_res_index = np.where(diff_res == res)[0][0] 
+    for res in diff_res_sorted[1:]:  # Loop through the rest of the differences
+        curr_res_index = np.where(diff_res == res)[0][0]
 
         index_diff = inf_point_list[curr_max_index] - inf_point_list[curr_res_index]
 
         # if difference between the max and the current is negligible and the index of the current is earlier then..
-        if ini_max_res - res < 0.05 and ( index_diff <= 3 and index_diff > 0 ):
-            curr_max_index = curr_res_index # replace the selected index with the current one
-
-    cutoff = inf_point_list[curr_max_index] + 5 # cutoff is the inflection point index + 5
-    primer = cons_seq_list[curr_max_index] # grab the correct consensus sequence as primer
+        if ini_max_res - res < 0.05 and (index_diff <= 3 and index_diff > 0):
+            curr_max_index = (
+                curr_res_index  # replace the selected index with the current one
+            )
+
+    cutoff = (
+        inf_point_list[curr_max_index] + 5
+    )  # cutoff is the inflection point index + 5
+    primer = cons_seq_list[
+        curr_max_index
+    ]  # grab the correct consensus sequence as primer
 
     # if the requested strand is reverse..
     if rev:
-        primer = str(Seq(primer).complement()) # ..get the complement of consensus sequence
+        primer = str(
+            Seq(primer).complement()
+        )  # ..get the complement of consensus sequence
 
     return cutoff, primer
 
+
 def main():
 
-    _PATH, _POINTS, _SAMPLE, _OUTPUT = parse_args()
-    inf_df = pd.read_csv(_POINTS, sep='\t')
+    path, points, sample, output = parse_args()
+    inf_df = pd.read_csv(points, sep="\t")
 
-    f_slice = inf_df[inf_df.strand == 'F'] # get forward inflection points
-    r_slice = inf_df[inf_df.strand == 'R'] # get reverse inflection points
+    f_slice = inf_df[inf_df.strand == "F"]  # get forward inflection points
+    r_slice = inf_df[inf_df.strand == "R"]  # get reverse inflection points
     r_slice = r_slice.reset_index(drop=True)
 
-    f_cutoff = ''
-    r_cutoff = ''
-    f_primer = ''
-    r_primer = ''
+    f_cutoff = ""
+    r_cutoff = ""
+    f_primer = ""
+    r_primer = ""
 
-    if not f_slice.empty: # if there is a forward inflection point..
+    if not f_slice.empty:  # if there is a forward inflection point..
         inf_list = f_slice.inf_point.tolist()
-        f_cutoff, f_primer = assess_inflection_point_mcp_for_sample(_PATH, inf_list) # .. assess and select
+        f_cutoff, f_primer = assess_inflection_point_mcp_for_sample(
+            path, inf_list
+        )  # .. assess and select
 
-    if not r_slice.empty: # if there is a reverse inflection point..
+    if not r_slice.empty:  # if there is a reverse inflection point..
         inf_list = r_slice.inf_point.tolist()
-        r_cutoff, r_primer = assess_inflection_point_mcp_for_sample(_PATH, inf_list, rev=True) # .. assess and select
+        r_cutoff, r_primer = assess_inflection_point_mcp_for_sample(
+            path, inf_list, rev=True
+        )  # .. assess and select
 
     # Output cutoff point(s) to .txt file
-    with open(f'{_OUTPUT}/{_SAMPLE}_cutoff.txt', 'w') as fw:
-        if f_cutoff != '':
-            fw.write(f'F: {f_cutoff}\n')
-        if r_cutoff != '':
-            fw.write(f'R: {r_cutoff}\n')
+    with open(f"{output}/{sample}_cutoff.txt", "w") as fw:
+        if f_cutoff != "":
+            fw.write(f"F: {f_cutoff}\n")
+        if r_cutoff != "":
+            fw.write(f"R: {r_cutoff}\n")
 
     # Output consensus primer sequence(s) to .fasta file
-    with open(f'{_OUTPUT}/{_SAMPLE}_auto_primers.fasta', 'w') as fw:
-        if f_cutoff != '':
-            fw.write(f'>F_auto\n{f_primer}\n')
-        if r_cutoff != '':
-            fw.write(f'>R_auto\n{r_primer}\n')
-
+    with open(f"{output}/{sample}_auto_primers.fasta", "w") as fw:
+        if f_cutoff != "":
+            fw.write(f">F_auto\n{f_primer}\n")
+        if r_cutoff != "":
+            fw.write(f">R_auto\n{r_primer}\n")
 
 
 if __name__ == "__main__":
-    main()
+    main()

mgnify_pipelines_toolkit/analysis/amplicon/assess_mcp_proportions.py

@@ -16,33 +16,51 @@
 
 import argparse
 from collections import defaultdict
-import subprocess
 
 import pandas as pd
 import numpy as np
 
-from mgnify_pipelines_toolkit.analysis.amplicon.amplicon_utils import get_read_count, build_cons_seq, build_mcp_cons_dict_list, fetch_mcp
+from mgnify_pipelines_toolkit.analysis.amplicon.amplicon_utils import (
+    get_read_count,
+    build_cons_seq,
+    build_mcp_cons_dict_list,
+    fetch_mcp,
+)
 from mgnify_pipelines_toolkit.constants.thresholds import MCP_MAX_LINE_COUNT
 
+
 def parse_args():
 
     parser = argparse.ArgumentParser()
 
-    parser.add_argument("-i", "--input", required=True, type=str, help="Path to fastq file to assess mcps")
+    parser.add_argument(
+        "-i",
+        "--input",
+        required=True,
+        type=str,
+        help="Path to fastq file to assess mcps",
+    )
     parser.add_argument("-s", "--sample", required=True, type=str, help="Sample ID")
-    parser.add_argument("-st", "--strand", required=True, choices=['FR', 'F', 'R'], help='F: Forward, R: Reverse')
+    parser.add_argument(
+        "-st",
+        "--strand",
+        required=True,
+        choices=["FR", "F", "R"],
+        help="F: Forward, R: Reverse",
+    )
     parser.add_argument("-o", "--output", required=True, type=str, help="Output path")
 
     args = parser.parse_args()
-    
-    _PATH = args.input
-    _SAMPLE = args.sample
-    _STRAND = args.strand
-    _OUTPUT = args.output
 
-    return _PATH, _SAMPLE, _STRAND, _OUTPUT
+    path = args.input
+    sample = args.sample
+    strand = args.strand
+    output = args.output
+
+    return path, sample, strand, output
+
 
-def find_mcp_props_for_sample(_PATH, rev=False):
+def find_mcp_props_for_sample(path, rev=False):
     """
     Generate mcp proportions in a stepwise and windowed manner for a fastq file.
 
@@ -50,38 +68,47 @@ def find_mcp_props_for_sample(_PATH, rev=False):
     Calculate the average conservation of the most common base at each index of a window.
     The resulting list of mcp conservations can be considered a conservation curve and used to
     identify inflection points where the conservation suddenly changes.
-    
+
     Output a dictionary where:
         key -> an index starting point e.g. base 10
         val -> the average conservation of the most common base for the mcp window goign from base 10 to 15 (inclusive)
     """
 
     res_dict = defaultdict(float)
-    start_range = range(2, 25, 1) # Range of starting indices
-    
-    print(f'Processing {_PATH}')
+    start_range = range(2, 25, 1)  # Range of starting indices
+
+    print(f"Processing {path}")
 
-    mcp_len = 5 # length of generated mcps
+    mcp_len = 5  # length of generated mcps
 
     for start in start_range:
 
-        end = start+mcp_len-1 # compute the final index for the mcp (inclusive). Indices are of base 1 not 0.
+        end = (
+            start + mcp_len - 1
+        )  # compute the final index for the mcp (inclusive). Indices are of base 1 not 0.
+
+        read_count = get_read_count(path, type="fastq")  # get read count for fastq file
 
-        read_count = get_read_count(_PATH, type='fastq') # get read count for fastq file
-        
         max_line_count = None
         if read_count > MCP_MAX_LINE_COUNT:
             max_line_count = MCP_MAX_LINE_COUNT
 
-        mcp_count_dict = fetch_mcp(_PATH, end, start, rev, max_line_count) # get MCP count dict
-        mcp_cons_list = build_mcp_cons_dict_list(mcp_count_dict, mcp_len) # list of base conservation dicts for mcps
-        cons_seq, cons_conf = build_cons_seq(mcp_cons_list, read_count, max_line_count=max_line_count) # get list of max base conservations for each index
-        
-        res_dict[start] = np.mean(cons_conf) # compute the mean
+        mcp_count_dict = fetch_mcp(
+            path, end, start, rev, max_line_count
+        )  # get MCP count dict
+        mcp_cons_list = build_mcp_cons_dict_list(
+            mcp_count_dict, mcp_len
+        )  # list of base conservation dicts for mcps
+        cons_seq, cons_conf = build_cons_seq(
+            mcp_cons_list, read_count, max_line_count=max_line_count
+        )  # get list of max base conservations for each index
+
+        res_dict[start] = np.mean(cons_conf)  # compute the mean
 
     return res_dict
 
-def concat_out(fwd_out='', rev_out=''):
+
+def concat_out(fwd_out="", rev_out=""):
     """
     Generate Pandas dataframe out of mcp dictionary.
 
@@ -97,16 +124,16 @@ def concat_out(fwd_out='', rev_out=''):
     df_ind = []
 
     # Check if fwd strand was requested
-    if fwd_out != '':
-        [ total_res_dict[key].append(fwd_out[key]) for key in fwd_out.keys() ]
-        df_ind.append('F')
+    if fwd_out != "":
+        [total_res_dict[key].append(fwd_out[key]) for key in fwd_out.keys()]
+        df_ind.append("F")
 
     # Check if rev strand was requested
-    if rev_out != '':
-        [ total_res_dict[key].append(rev_out[key]) for key in rev_out.keys() ] 
-        df_ind.append('R')
+    if rev_out != "":
+        [total_res_dict[key].append(rev_out[key]) for key in rev_out.keys()]
+        df_ind.append("R")
 
-    res_df= pd.DataFrame.from_dict(total_res_dict)
+    res_df = pd.DataFrame.from_dict(total_res_dict)
     res_df.index = df_ind
 
     return res_df
@@ -114,31 +141,33 @@ def concat_out(fwd_out='', rev_out=''):
 
 def main():
 
-    _PATH, _SAMPLE, _STRAND, _OUTPUT = parse_args()
+    path, sample, strand, output = parse_args()
 
-    res_df = ''
+    res_df = ""
 
     # TODO: match-case statement is python 3.10>. We are currently locking the version
     # at version 3.9. The day we bump the version we should replace these if statements
     # with a match-case block.
 
-    if _STRAND == "FR":
-        fwd_out = find_mcp_props_for_sample(_PATH)
-        rev_out = find_mcp_props_for_sample(_PATH, rev=True)
+    if strand == "FR":
+        fwd_out = find_mcp_props_for_sample(path)
+        rev_out = find_mcp_props_for_sample(path, rev=True)
         res_df = concat_out(fwd_out, rev_out)
-    elif _STRAND == "F":
-        fwd_out = find_mcp_props_for_sample(_PATH)
+    elif strand == "F":
+        fwd_out = find_mcp_props_for_sample(path)
         res_df = concat_out(fwd_out)
-    elif _STRAND == "R":
-        rev_out = find_mcp_props_for_sample(_PATH, rev=True)
+    elif strand == "R":
+        rev_out = find_mcp_props_for_sample(path, rev=True)
         res_df = concat_out(rev_out=rev_out)
     else:
-        print("Incorrect strand input. Should be F for forward, R for reverse, or FR for both.")
+        print(
+            "Incorrect strand input. Should be F for forward, R for reverse, or FR for both."
+        )
         exit(1)
 
     # Save resulting dataframe to a tsv file
-    res_df.to_csv(f'{_OUTPUT}/{_SAMPLE}_mcp_cons.tsv', sep='\t')
+    res_df.to_csv(f"{output}/{sample}_mcp_cons.tsv", sep="\t")
+
 
-    
 if __name__ == "__main__":
-    main()
+    main()