More cleanup

.github/workflows/build.yml

@@ -10,7 +10,7 @@ jobs:
     - name: Install dependencies
       run: poetry install
     - name: Build figures
-      run: make -j 2 all
+      run: make -i all
     - name: Upload files
       uses: actions/upload-artifact@v4
       with:

ddmc/clustering.py

@@ -70,13 +70,11 @@ def _m_step(self, X: np.ndarray, log_resp: np.ndarray):
         """
         if self._missing:
             labels = np.argmax(log_resp, axis=1)
-            centers = self.means_.T  # samples x clusters
+            centers = np.array(self.means_)  # samples x clusters
+            centers_fill = centers[labels, :]
 
-            assert len(labels) == X.shape[0]
-            for ii in range(X.shape[0]):  # X is peptides x samples
-                X[ii, self.missing_d[ii, :]] = centers[
-                    self.missing_d[ii, :], labels[ii]
-                ]
+            assert centers_fill.shape == X.shape
+            X[self.missing_d] = centers_fill[self.missing_d]
 
         super()._m_step(X, log_resp)  # Do the regular m step
 

ddmc/figures/common.py

@@ -126,7 +126,7 @@ def genFigure():
     print(f"Figure {sys.argv[1]} is done after {time.time() - start} seconds.\n")
 
 
-def getDDMC_CPTAC(n_components: int, SeqWeight: float) -> DDMC:
+def getDDMC_CPTAC(n_components: int, SeqWeight: float):
     # Import signaling data
     X = filter_NaNpeptides(
         pd.read_csv("ddmc/data/MS/CPTAC/CPTAC-preprocessedMotfis.csv").iloc[:, 1:],
@@ -143,7 +143,7 @@ def getDDMC_CPTAC(n_components: int, SeqWeight: float) -> DDMC:
         distance_method="Binomial",
         random_state=5,
     ).fit(d)
-    return model
+    return model, X
 
 
 def plotMotifs(pssm, ax: axes.Axes, titles=False, yaxis=False):

ddmc/figures/figureM4.py

@@ -46,9 +46,9 @@ def makeFigure():
     ax[0].legend(prop={"size": 5}, loc=0)
 
     # Fit Data, Mix, and Seq Models
-    dataM = getDDMC_CPTAC(n_components=30, SeqWeight=0.0)
-    mixM = getDDMC_CPTAC(n_components=30, SeqWeight=250.0)
-    seqM = getDDMC_CPTAC(n_components=30, SeqWeight=1.0e6)
+    dataM, _ = getDDMC_CPTAC(n_components=30, SeqWeight=0.0)
+    mixM, _ = getDDMC_CPTAC(n_components=30, SeqWeight=250.0)
+    seqM, _ = getDDMC_CPTAC(n_components=30, SeqWeight=1.0e6)
     models = [dataM, mixM, seqM]
 
     # Center to peptide distance

ddmc/figures/figureM5.py

@@ -24,14 +24,8 @@ def makeFigure():
     # Get list of axis objects
     ax, f = getSetup((11, 10), (3, 3), multz={0: 1, 4: 1})
 
-    # Import signaling data
-    X = filter_NaNpeptides(
-        pd.read_csv("ddmc/data/MS/CPTAC/CPTAC-preprocessedMotfis.csv").iloc[:, 1:],
-        tmt=2,
-    )
-
     # Fit DDMC
-    model = getDDMC_CPTAC(n_components=30, SeqWeight=100.0)
+    model, X = getDDMC_CPTAC(n_components=30, SeqWeight=100.0)
 
     # Normalize
     centers = pd.DataFrame(model.transform()).T

ddmc/figures/figureM6.py

@@ -26,14 +26,8 @@ def makeFigure():
     # Get list of axis objects
     ax, f = getSetup((11, 7), (2, 3), multz={0: 1})
 
-    # Import signaling data
-    X = filter_NaNpeptides(
-        pd.read_csv("ddmc/data/MS/CPTAC/CPTAC-preprocessedMotfis.csv").iloc[:, 1:],
-        tmt=2,
-    )
-
     # Fit DDMC
-    model = getDDMC_CPTAC(n_components=30, SeqWeight=100.0)
+    model, X = getDDMC_CPTAC(n_components=30, SeqWeight=100.0)
 
     # Import Genotype data
     mutations = pd.read_csv("ddmc/data/MS/CPTAC/Patient_Mutations.csv")

ddmc/figures/figureMS2.py

@@ -2,38 +2,25 @@
 This creates Supplemental Figure 2: Cluster motifs
 """
 
-import pandas as pd
 import numpy as np
-from .common import getSetup
+from .common import getSetup, getDDMC_CPTAC
 from .common import plotMotifs
-from ..pre_processing import filter_NaNpeptides
-from ..clustering import DDMC
 
 
 def makeFigure():
     """Get a list of the axis objects and create a figure"""
     # Get list of axis objects
     ax, f = getSetup((9, 9), (5, 5))
 
-    # Import signaling data
-    X = filter_NaNpeptides(
-        pd.read_csv("ddmc/data/MS/CPTAC/CPTAC-preprocessedMotfis.csv").iloc[:, 1:],
-        tmt=2,
-    )
-    d = X.select_dtypes(include=[float]).T
-    i = X["Sequence"]
-
     # Fit DDMC
-    model = DDMC(
-        i, n_components=30, SeqWeight=100, distance_method="Binomial", random_state=5
-    ).fit(d)
+    model, _ = getDDMC_CPTAC(n_components=30, SeqWeight=100.0)
 
     pssms, cl_num = model.pssms(PsP_background=False)
     ylabels = np.arange(0, 21, 5)
     xlabels = [20, 21, 22, 23, 24, 25]
     for ii, cc in enumerate(cl_num):
         cluster = "Cluster " + str(cc)
-        plotMotifs(pssms[ii], axes=ax[ii], titles=cluster, yaxis=[0, 10])
+        plotMotifs(pssms[ii], ax=ax[ii], titles=cluster, yaxis=[0, 10])
         if ii not in ylabels:
             ax[ii].set_ylabel("")
             ax[ii].get_yaxis().set_visible(False)

ddmc/figures/figureMS3.py

@@ -4,10 +4,8 @@
 
 import pandas as pd
 from sklearn.linear_model import LogisticRegressionCV
-from .common import getSetup
+from .common import getSetup, getDDMC_CPTAC
 from .figureM4 import TransformCenters, HotColdBehavior, find_patients_with_NATandTumor
-from ..pre_processing import filter_NaNpeptides
-from ..clustering import DDMC
 from ..logistic_regression import plotROC
 
 
@@ -16,14 +14,6 @@ def makeFigure():
     # Get list of axis objects
     ax, f = getSetup((15, 10), (3, 5))
 
-    # Signaling
-    X = filter_NaNpeptides(
-        pd.read_csv("ddmc/data/MS/CPTAC/CPTAC-preprocessedMotfis.csv").iloc[:, 1:],
-        tmt=2,
-    )
-    d = X.select_dtypes(include=[float]).T
-    i = X["Sequence"]
-
     # Genotype data
     mutations = pd.read_csv("ddmc/data/MS/CPTAC/Patient_Mutations.csv")
     mOI = mutations[
@@ -46,7 +36,7 @@ def makeFigure():
     folds = 5
     weights = [0, 100, 500, 1000, 1000000]
     for ii, w in enumerate(weights):
-        model = DDMC(i, n_components=30, SeqWeight=w, distance_method="Binomial").fit(d)
+        model, X = getDDMC_CPTAC(n_components=30, SeqWeight=w)
 
         # Find and scale centers
         centers_gen, centers_hcb = TransformCenters(model, X)
@@ -63,7 +53,7 @@ def makeFigure():
             ax[ii],
             lr,
             centers_gen.values,
-            y["STK11.mutation.status"],
+            y["STK11.mutation.status"].values, # type: ignore
             cv_folds=folds,
             title="STK11m " + "w=" + str(model.SeqWeight) + prio,
         )
@@ -73,7 +63,7 @@ def makeFigure():
             ax[ii + 5],
             lr,
             centers_gen.values,
-            y["EGFR.mutation.status"],
+            y["EGFR.mutation.status"].values, # type: ignore
             cv_folds=folds,
             title="EGFRm " + "w=" + str(model.SeqWeight) + prio,
         )
@@ -84,7 +74,7 @@ def makeFigure():
             ax[ii + 10],
             lr,
             centers_hcb.values,
-            y_hcb,
+            y_hcb.values,
             cv_folds=folds,
             title="Infiltration " + "w=" + str(model.SeqWeight) + prio,
         )

ddmc/figures/figureMS6.py

@@ -22,7 +22,7 @@ def makeFigure():
     ax, f = getSetup((11, 7), (2, 3), multz={0: 1})
 
     # Fit DDMC
-    model = getDDMC_CPTAC(n_components=30, SeqWeight=100.0)
+    model, X = getDDMC_CPTAC(n_components=30, SeqWeight=100.0)
 
     # Import Genotype data
     mutations = pd.read_csv("ddmc/data/MS/CPTAC/Patient_Mutations.csv")