More cleanup

ddmc/figures/figureM3.py

@@ -10,7 +10,7 @@
 from ..validations import preprocess_ebdt_mcf7
 from .common import subplotLabel, getSetup
 from ..pca import plotPCA
-from .figure2 import plotDistanceToUpstreamKinase, plotMotifs, ShuffleClusters
+from .common import plotDistanceToUpstreamKinase, plotMotifs, ShuffleClusters
 from .figureM5 import plot_NetPhoresScoreByKinGroup
 from ..clustering import compute_control_pssm
 from ..binomial import AAlist

ddmc/figures/figureM5.py

@@ -16,7 +16,7 @@
 from ..clustering import DDMC
 from .common import subplotLabel, getSetup
 from ..logistic_regression import plotClusterCoefficients, plotROC
-from ..figures.figure2 import plotDistanceToUpstreamKinase
+from .common import plotDistanceToUpstreamKinase
 from ..pca import plotPCA
 from ..pre_processing import filter_NaNpeptides
 

ddmc/figures/figureM6.py

@@ -13,7 +13,7 @@
 from bioinfokit import visuz
 from ..clustering import DDMC
 from ..pre_processing import filter_NaNpeptides
-from .figure2 import plotDistanceToUpstreamKinase
+from .common import plotDistanceToUpstreamKinase
 from .figureM4 import find_patients_with_NATandTumor
 from .figureM5 import (
     plot_clusters_binaryfeatures,

ddmc/figures/figureM7.py

@@ -15,7 +15,7 @@
     calculate_mannW_pvals,
     plot_clusters_binaryfeatures,
 )
-from .figure2 import plotDistanceToUpstreamKinase
+from .commmon import plotDistanceToUpstreamKinase
 from ..clustering import DDMC
 from ..logistic_regression import plotROC, plotClusterCoefficients
 from ..pre_processing import filter_NaNpeptides

ddmc/figures/figureMS2.py

@@ -7,7 +7,7 @@
 import numpy as np
 import seaborn as sns
 from .common import subplotLabel, getSetup
-from .figure2 import plotMotifs
+from .common import plotMotifs
 from ..pre_processing import filter_NaNpeptides
 from ..clustering import DDMC
 

ddmc/figures/figureMS6.py

@@ -9,7 +9,7 @@
 from sklearn.preprocessing import StandardScaler
 from ..clustering import DDMC
 from ..pre_processing import filter_NaNpeptides
-from .figure2 import plotDistanceToUpstreamKinase
+from .common import plotDistanceToUpstreamKinase
 from .figureM4 import find_patients_with_NATandTumor
 from .figureM5 import (
     plot_clusters_binaryfeatures,

ddmc/pre_processing.py

@@ -12,115 +12,6 @@
 
 
 ###-------------------------- Pre-processing MS data --------------------------###
-def preprocessing(
-    AXLwt_GF=False,
-    AXLm_ErlAF154=False,
-    AXL_Das_DR=False,
-    Vfilter=False,
-    FCfilter=False,
-    log2T=False,
-    rawdata=False,
-    corrCut=0.5,
-):
-    """Input: Raw MS bio-replicates. Output: Mean-centered merged data set.
-    1. Concatenation, 2. log-2 transformation, 3. Mean-Center, 4. Merging, 5. Fold-change,
-    6. Filters: 'Vfilter' filters by correlation when 2 overlapping peptides or std cutoff if >= 3.
-    Note 1: 'motifs' redefines peptide sequences as XXXXXyXXXXX which affects merging.
-    Note 2: Data is converted back to linear scale before filtering so 'log2T=True' to use log-scale for analysis.
-    Note 3: CPTAC is already normalized, so: mc_row and mc_col = False"""
-    filesin = list()
-
-    if AXLwt_GF:
-        filesin.append(
-            pd.read_csv(
-                os.path.join(
-                    path,
-                    "./data/MS/GrowthFactors/20180817_JG_AM_TMT10plex_R1_psms_raw.csv",
-                )
-            )
-        )
-        filesin.append(
-            pd.read_csv(
-                os.path.join(
-                    path,
-                    "./data/MS/GrowthFactors/20190214_JG_AM_PC9_AXL_TMT10_AC28_R2_PSMs_raw.csv",
-                )
-            )
-        )
-        filesin.append(
-            pd.read_csv(
-                os.path.join(path, "./data/MS/GrowthFactors/CombinedBR3_TR1&2_raw.csv")
-            )
-        )
-    if AXLm_ErlAF154:
-        filesin.append(
-            pd.read_csv(
-                os.path.join(
-                    path, "./data/MS/AXL/PC9_mutants_ActivatingAb_BR1_raw_wAcc.csv"
-                )
-            )
-        )
-        filesin.append(
-            pd.read_csv(
-                os.path.join(
-                    path, "./data/MS/AXL/PC9_mutants_ActivatingAb_BR3_raw_wAcc.csv"
-                )
-            )
-        )
-        filesin.append(
-            pd.read_csv(
-                os.path.join(
-                    path, "./data/MS/AXL/PC9_mutants_ActivatingAb_BR4_raw_wAcc.csv"
-                )
-            )
-        )
-    if AXL_Das_DR:
-        filesin.append(
-            pd.read_csv(
-                "ddmc/data/Validations/Experimental/MassSpec/06232021-DasDR_BR1_raw.csv"
-            ).iloc[:, 1:]
-        )
-        filesin.append(
-            pd.read_csv(
-                "ddmc/data/Validations/Experimental/MassSpec/06232021-DasDR_BR2_raw.csv"
-            ).iloc[:, 1:]
-        )
-
-    data_headers = list(filesin[0].select_dtypes(include=["float64"]).columns)
-    FCto = data_headers[0]
-    X = Log2T(pd.concat(filesin))
-
-    # mean center rows
-    numeric_cols = X.select_dtypes(include=["float64"]).columns
-    X[numeric_cols] = X[numeric_cols].sub(X[numeric_cols].mean(axis=1), axis=0)
-
-    fullnames, genes = FormatName(X)
-    X["Protein"] = fullnames
-    X.insert(3, "Gene", genes)
-    merging_indices = list(X.select_dtypes(include=["object"]).columns)
-
-    if rawdata:
-        return X
-
-    X = MapMotifs(X, genes)
-    merging_indices.insert(3, "Position")
-
-    if Vfilter:
-        X = VFilter(X, merging_indices, data_headers, corrCut=corrCut, stdCut=0.6)
-
-    X = MergeDfbyMean(X.copy(), data_headers, merging_indices).reset_index()[
-        merging_indices + data_headers
-    ]
-
-    if FCfilter:
-        X = FoldChangeFilterBasedOnMaxFC(X, data_headers, cutoff=0.40)
-
-    if not log2T:
-        X = Linear(X, data_headers)
-
-    return X
-
-
 def preprocessCPTAC():
     """Replace patient identifiers, fill NaNs, and make it compatible with current code."""
     X = pd.read_csv(
@@ -352,6 +243,16 @@ def TripsMeanAndStd(triplicates, merging_indices, data_headers):
     return X.reset_index()
 
 
+def MeanCenter(X, mc_row, mc_col):
+    """ Mean centers each row of values. logT also optionally log2-transforms. """
+    data_headers = X.select_dtypes(include=["float64"]).columns
+    if mc_row:
+        X[data_headers] = X[data_headers].sub(X[data_headers].mean(axis=1), axis=0)
+    if mc_col:
+        X[data_headers] = X[data_headers].sub(X[data_headers].mean(axis=0), axis=1)
+    return X
+
+
 def FilterByRange(X, rangeCut=0.4):
     """Filter rows for those containing more than a range threshold."""
     Rg = X.iloc[:, X.columns.get_level_values(1) == "ptp"]

ddmc/tests/test_CoClustering.py

@@ -4,11 +4,13 @@
 
 import pytest
 import numpy as np
+import pandas as pd
 from ..clustering import DDMC
-from ..pre_processing import preprocessing
+from ..pre_processing import filter_NaNpeptides
 
 
-X = preprocessing(AXLwt_GF=True, Vfilter=True, FCfilter=True, log2T=True)
+X = pd.read_csv("ddmc/data/MS/CPTAC/CPTAC-preprocessedMotfis.csv").iloc[:, 1:]
+X = filter_NaNpeptides(X, tmt=25)
 data = X.select_dtypes(include=["float64"]).T
 info = X.select_dtypes(include=["object"])