Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,106 +1,96 @@
|
|
| 1 |
-
|
| 2 |
-
#import sys, subprocess
|
| 3 |
-
#def _need(modname):
|
| 4 |
-
# try: __import__(modname); return False
|
| 5 |
-
# except ImportError: return True
|
| 6 |
-
#def _pip(pkg_spec):
|
| 7 |
-
# subprocess.check_call([sys.executable, "-m", "pip", "install", pkg_spec])
|
| 8 |
-
|
| 9 |
-
#PKGS = [
|
| 10 |
-
# ("joblib", "1.4.2", "joblib"),
|
| 11 |
-
# ("numpy", "1.26.4","numpy"),
|
| 12 |
-
# ("pandas", "2.2.2", "pandas"),
|
| 13 |
-
# ("scikit-learn", "1.4.2", "sklearn"),
|
| 14 |
-
# ("scipy", "1.11.4","scipy"),
|
| 15 |
-
# ("shap", "0.46.0","shap"),
|
| 16 |
-
# ("matplotlib", "3.8.4", "matplotlib"),
|
| 17 |
-
# ("openpyxl", "3.1.5", "openpyxl"),
|
| 18 |
-
# ("pillow", None, "PIL"),
|
| 19 |
-
# ("gradio", "4.44.1","gradio"),
|
| 20 |
-
#]
|
| 21 |
-
#for pkg, ver, mod in PKGS:
|
| 22 |
-
# if _need(mod):
|
| 23 |
-
# _pip(f"{pkg}=={ver}" if ver else pkg)
|
| 24 |
-
# now safe to import everything
|
| 25 |
-
import os, json, io, gzip, joblib, traceback
|
| 26 |
import numpy as np, pandas as pd
|
| 27 |
-
|
|
|
|
|
|
|
| 28 |
import matplotlib.pyplot as plt
|
| 29 |
-
|
| 30 |
import shap
|
| 31 |
import gradio as gr
|
| 32 |
|
| 33 |
-
#
|
| 34 |
-
|
|
|
|
|
|
|
| 35 |
FEATS_PATH = "feature_names.json"
|
| 36 |
|
| 37 |
-
# if you uploaded a .gz model, uncomment these two lines and delete the next joblib.load:
|
| 38 |
-
# with gzip.open(MODEL_PATH, "rb") as f:
|
| 39 |
-
# model = joblib.load(f)
|
| 40 |
-
model = joblib.load(MODEL_PATH)
|
| 41 |
-
|
| 42 |
-
with open(FEATS_PATH) as f:
|
| 43 |
-
FEATURES = json.load(f)
|
| 44 |
-
|
| 45 |
LABEL_NAME = {0: "Benign", 1: "Pathogenic"}
|
| 46 |
THRESHOLD = 0.7
|
| 47 |
-
|
| 48 |
ID_CANDIDATES = [
|
| 49 |
"#Uploaded_Variation", "#Uploaded_variation", "uploaded_variation",
|
| 50 |
"Uploaded_Variation", "Uploaded_variation", "variant_id",
|
| 51 |
"Variant_ID", "variation_id", "Variant", "ID", "HGVSc", "HGVSp"
|
| 52 |
]
|
| 53 |
|
| 54 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
def _clean_headers(df):
|
| 56 |
-
df.columns = (
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
| 59 |
return df
|
| 60 |
|
| 61 |
def _read_table(file):
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
else:
|
| 66 |
-
# NOTE: pandas python engine does not support low_memory
|
| 67 |
try:
|
| 68 |
df = pd.read_csv(
|
| 69 |
-
|
| 70 |
engine="python",
|
| 71 |
-
sep=None,
|
| 72 |
dtype=str,
|
| 73 |
-
na_values=["", "NA", "N/A"],
|
| 74 |
quoting=3,
|
| 75 |
on_bad_lines="skip",
|
| 76 |
-
encoding_errors="ignore"
|
| 77 |
)
|
| 78 |
except Exception:
|
| 79 |
df = None
|
| 80 |
for sep in [",", "\t", ";", "|"]:
|
| 81 |
try:
|
| 82 |
df = pd.read_csv(
|
| 83 |
-
|
| 84 |
engine="python",
|
| 85 |
sep=sep,
|
| 86 |
dtype=str,
|
| 87 |
-
na_values=["", "NA", "N/A"],
|
| 88 |
quoting=3,
|
| 89 |
on_bad_lines="skip",
|
| 90 |
-
encoding_errors="ignore"
|
| 91 |
)
|
| 92 |
break
|
| 93 |
except Exception:
|
| 94 |
pass
|
| 95 |
if df is None:
|
| 96 |
raise
|
| 97 |
-
return _clean_headers(df)
|
| 98 |
|
|
|
|
| 99 |
|
| 100 |
def _coerce_numeric(df):
|
| 101 |
for c in df.columns:
|
| 102 |
-
|
| 103 |
-
df[c] = pd.to_numeric(df[c], errors="coerce")
|
| 104 |
return df
|
| 105 |
|
| 106 |
def _prepare(df):
|
|
@@ -115,119 +105,136 @@ def _validate_columns(df):
|
|
| 115 |
return missing, extra
|
| 116 |
|
| 117 |
def _pick_id_series(df):
|
| 118 |
-
# case-insensitive lookup without altering original headers
|
| 119 |
norm = {c.lower(): c for c in df.columns}
|
| 120 |
for cand in ID_CANDIDATES:
|
| 121 |
-
|
| 122 |
-
|
| 123 |
-
col = norm[key]
|
| 124 |
return df[col].astype(str), col
|
| 125 |
-
# fallback
|
| 126 |
return pd.Series(df.index, index=df.index, name="row_index"), "row_index"
|
| 127 |
|
| 128 |
-
#
|
|
|
|
|
|
|
| 129 |
def predict_csv(file):
|
| 130 |
try:
|
| 131 |
-
if file is None: return None, "No file."
|
| 132 |
df = _read_table(file)
|
| 133 |
missing, _ = _validate_columns(df)
|
| 134 |
-
if missing:
|
|
|
|
|
|
|
| 135 |
X = _prepare(df)
|
| 136 |
|
| 137 |
proba = model.predict_proba(X)[:, 1]
|
| 138 |
pred = (proba > THRESHOLD).astype(int)
|
| 139 |
|
| 140 |
id_series, id_name = _pick_id_series(df)
|
| 141 |
-
id_label = "Variant ID" if id_name != "row_index" else "Variant ID (
|
| 142 |
|
| 143 |
out = pd.DataFrame({
|
| 144 |
id_label: id_series.values,
|
| 145 |
"Predicted probability (pathogenic)": proba,
|
| 146 |
-
"Predicted class": np.where(pred == 1,
|
| 147 |
})
|
| 148 |
|
| 149 |
n = len(out)
|
| 150 |
n_path = int((out["Predicted class"] == "Pathogenic").sum())
|
| 151 |
-
|
| 152 |
txt = (
|
| 153 |
f"Processed {n} records at threshold {THRESHOLD:.2f}. "
|
| 154 |
-
f"Pathogenic: {n_path} ({(n_path/n*100):.1f}%).
|
| 155 |
-
|
| 156 |
-
"
|
| 157 |
)
|
|
|
|
| 158 |
return out, txt
|
|
|
|
| 159 |
except Exception:
|
| 160 |
-
return None, "ERROR:\n" + traceback.format_exc()
|
| 161 |
|
|
|
|
|
|
|
|
|
|
| 162 |
def explain_first_row(file):
|
| 163 |
try:
|
| 164 |
-
if file is None: return None, "No file."
|
| 165 |
df = _read_table(file)
|
| 166 |
missing, _ = _validate_columns(df)
|
| 167 |
-
if missing:
|
|
|
|
|
|
|
| 168 |
X = _prepare(df)
|
| 169 |
-
if X
|
|
|
|
| 170 |
|
| 171 |
-
|
| 172 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 173 |
|
| 174 |
-
|
| 175 |
-
|
| 176 |
-
ex = expl(X_imp.iloc[[0]])
|
| 177 |
-
vals = np.array(ex.values)
|
| 178 |
vals_1d = vals[0, :, 1] if vals.ndim == 3 else vals[0]
|
| 179 |
-
s = pd.Series(vals_1d, index=FEATURES).sort_values(key=np.abs, ascending=False).head(15)
|
| 180 |
|
| 181 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 182 |
ax = fig.add_subplot(111)
|
| 183 |
-
s.iloc[::-1].plot(kind="barh", ax=ax
|
| 184 |
ax.set_title("SHAP values (Sample #1)")
|
| 185 |
fig.tight_layout()
|
| 186 |
|
| 187 |
-
# clinical note: map top features to ACMG evidence “families”
|
| 188 |
-
feats = [f.lower() for f in s.index.tolist()]
|
| 189 |
-
acmg_tags = []
|
| 190 |
-
if any("af" in f or "gnomad" in f or "freq" in f for f in feats):
|
| 191 |
-
acmg_tags.append("Population frequency → BA1/BS1 (benign thresholds) context")
|
| 192 |
-
if any("score" in f or "rankscore" in f or "meta" in f or "deleterious" in f for f in feats):
|
| 193 |
-
acmg_tags.append("Computational evidence → PP3 (supporting pathogenic) or BP4 (supporting benign)")
|
| 194 |
-
if any("domain" in f or "hotspot" in f or "pfam" in f for f in feats):
|
| 195 |
-
acmg_tags.append("Protein domain/hotspot → PM1 (moderate) if curated hotspot")
|
| 196 |
-
if any("blosum" in f or "aminoacid" in f or "aa" in f for f in feats):
|
| 197 |
-
acmg_tags.append("Same-residue considerations → PM5 possible with prior pathogenic missense")
|
| 198 |
-
|
| 199 |
-
tag_txt = " | ".join(acmg_tags) if acmg_tags else "Top features map to PP3/BP4 computational evidence."
|
| 200 |
note = (
|
| 201 |
-
"
|
| 202 |
-
|
| 203 |
-
"
|
| 204 |
)
|
|
|
|
| 205 |
return fig, note
|
|
|
|
| 206 |
except Exception:
|
| 207 |
-
return None, "ERROR:\n" + traceback.format_exc()
|
| 208 |
|
| 209 |
-
#
|
|
|
|
|
|
|
| 210 |
with gr.Blocks(title="Pathogenicity Predictor") as demo:
|
| 211 |
-
gr.HTML(
|
| 212 |
-
|
| 213 |
-
|
| 214 |
-
|
| 215 |
-
""")
|
| 216 |
-
|
| 217 |
|
|
|
|
| 218 |
|
| 219 |
-
|
| 220 |
-
inp = gr.File(label="Upload CSV/XLSX containing the trained feature columns")
|
| 221 |
with gr.Row():
|
| 222 |
btn_pred = gr.Button("Predict")
|
| 223 |
-
btn_expl = gr.Button("Explain Sample #1")
|
|
|
|
| 224 |
out_df = gr.Dataframe(label="Predictions")
|
| 225 |
-
out_plot = gr.Plot(label="SHAP
|
| 226 |
out_txt = gr.Markdown()
|
| 227 |
-
btn_pred.click(fn=predict_csv, inputs=inp, outputs=[out_df, out_txt])
|
| 228 |
-
btn_expl.click(fn=explain_first_row, inputs=inp, outputs=[out_plot, out_txt])
|
| 229 |
|
| 230 |
-
|
| 231 |
-
|
| 232 |
-
|
| 233 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import os, json, traceback, gzip, joblib
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
import numpy as np, pandas as pd
|
| 3 |
+
|
| 4 |
+
import matplotlib
|
| 5 |
+
matplotlib.use("Agg")
|
| 6 |
import matplotlib.pyplot as plt
|
| 7 |
+
|
| 8 |
import shap
|
| 9 |
import gradio as gr
|
| 10 |
|
| 11 |
+
# =========================
|
| 12 |
+
# Paths / constants
|
| 13 |
+
# =========================
|
| 14 |
+
MODEL_PATH = "model_extratrees_ds4.joblib" # or .gz
|
| 15 |
FEATS_PATH = "feature_names.json"
|
| 16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
LABEL_NAME = {0: "Benign", 1: "Pathogenic"}
|
| 18 |
THRESHOLD = 0.7
|
| 19 |
+
|
| 20 |
ID_CANDIDATES = [
|
| 21 |
"#Uploaded_Variation", "#Uploaded_variation", "uploaded_variation",
|
| 22 |
"Uploaded_Variation", "Uploaded_variation", "variant_id",
|
| 23 |
"Variant_ID", "variation_id", "Variant", "ID", "HGVSc", "HGVSp"
|
| 24 |
]
|
| 25 |
|
| 26 |
+
# =========================
|
| 27 |
+
# Load model
|
| 28 |
+
# =========================
|
| 29 |
+
if MODEL_PATH.endswith(".gz"):
|
| 30 |
+
with gzip.open(MODEL_PATH, "rb") as f:
|
| 31 |
+
model = joblib.load(f)
|
| 32 |
+
else:
|
| 33 |
+
model = joblib.load(MODEL_PATH)
|
| 34 |
+
|
| 35 |
+
with open(FEATS_PATH) as f:
|
| 36 |
+
FEATURES = json.load(f)
|
| 37 |
+
|
| 38 |
+
IS_PIPELINE = hasattr(model, "named_steps")
|
| 39 |
+
|
| 40 |
+
# =========================
|
| 41 |
+
# Helpers
|
| 42 |
+
# =========================
|
| 43 |
def _clean_headers(df):
|
| 44 |
+
df.columns = (
|
| 45 |
+
df.columns.astype(str)
|
| 46 |
+
.str.replace(r'[\u200b\ufeff]', '', regex=True)
|
| 47 |
+
.str.strip()
|
| 48 |
+
)
|
| 49 |
return df
|
| 50 |
|
| 51 |
def _read_table(file):
|
| 52 |
+
if file is None:
|
| 53 |
+
raise ValueError("No file provided")
|
| 54 |
+
|
| 55 |
+
path = file.name if hasattr(file, "name") else file
|
| 56 |
+
|
| 57 |
+
if path.lower().endswith((".xlsx", ".xls")):
|
| 58 |
+
df = pd.read_excel(path, engine="openpyxl")
|
| 59 |
else:
|
|
|
|
| 60 |
try:
|
| 61 |
df = pd.read_csv(
|
| 62 |
+
path,
|
| 63 |
engine="python",
|
| 64 |
+
sep=None,
|
| 65 |
dtype=str,
|
|
|
|
| 66 |
quoting=3,
|
| 67 |
on_bad_lines="skip",
|
| 68 |
+
encoding_errors="ignore"
|
| 69 |
)
|
| 70 |
except Exception:
|
| 71 |
df = None
|
| 72 |
for sep in [",", "\t", ";", "|"]:
|
| 73 |
try:
|
| 74 |
df = pd.read_csv(
|
| 75 |
+
path,
|
| 76 |
engine="python",
|
| 77 |
sep=sep,
|
| 78 |
dtype=str,
|
|
|
|
| 79 |
quoting=3,
|
| 80 |
on_bad_lines="skip",
|
| 81 |
+
encoding_errors="ignore"
|
| 82 |
)
|
| 83 |
break
|
| 84 |
except Exception:
|
| 85 |
pass
|
| 86 |
if df is None:
|
| 87 |
raise
|
|
|
|
| 88 |
|
| 89 |
+
return _clean_headers(df)
|
| 90 |
|
| 91 |
def _coerce_numeric(df):
|
| 92 |
for c in df.columns:
|
| 93 |
+
df[c] = pd.to_numeric(df[c], errors="coerce")
|
|
|
|
| 94 |
return df
|
| 95 |
|
| 96 |
def _prepare(df):
|
|
|
|
| 105 |
return missing, extra
|
| 106 |
|
| 107 |
def _pick_id_series(df):
|
|
|
|
| 108 |
norm = {c.lower(): c for c in df.columns}
|
| 109 |
for cand in ID_CANDIDATES:
|
| 110 |
+
if cand.lower() in norm:
|
| 111 |
+
col = norm[cand.lower()]
|
|
|
|
| 112 |
return df[col].astype(str), col
|
|
|
|
| 113 |
return pd.Series(df.index, index=df.index, name="row_index"), "row_index"
|
| 114 |
|
| 115 |
+
# =========================
|
| 116 |
+
# Prediction
|
| 117 |
+
# =========================
|
| 118 |
def predict_csv(file):
|
| 119 |
try:
|
|
|
|
| 120 |
df = _read_table(file)
|
| 121 |
missing, _ = _validate_columns(df)
|
| 122 |
+
if missing:
|
| 123 |
+
return None, f"Missing {len(missing)} features. First 10: {missing[:10]}"
|
| 124 |
+
|
| 125 |
X = _prepare(df)
|
| 126 |
|
| 127 |
proba = model.predict_proba(X)[:, 1]
|
| 128 |
pred = (proba > THRESHOLD).astype(int)
|
| 129 |
|
| 130 |
id_series, id_name = _pick_id_series(df)
|
| 131 |
+
id_label = "Variant ID" if id_name != "row_index" else "Variant ID (row index)"
|
| 132 |
|
| 133 |
out = pd.DataFrame({
|
| 134 |
id_label: id_series.values,
|
| 135 |
"Predicted probability (pathogenic)": proba,
|
| 136 |
+
"Predicted class": np.where(pred == 1, "Pathogenic", "Benign"),
|
| 137 |
})
|
| 138 |
|
| 139 |
n = len(out)
|
| 140 |
n_path = int((out["Predicted class"] == "Pathogenic").sum())
|
| 141 |
+
|
| 142 |
txt = (
|
| 143 |
f"Processed {n} records at threshold {THRESHOLD:.2f}. "
|
| 144 |
+
f"Pathogenic: {n_path} ({(n_path/n*100):.1f}%).\n\n"
|
| 145 |
+
"⚠️ Model output is *supportive only*. "
|
| 146 |
+
"Final classification must follow ACMG/AMP guidelines."
|
| 147 |
)
|
| 148 |
+
|
| 149 |
return out, txt
|
| 150 |
+
|
| 151 |
except Exception:
|
| 152 |
+
return None, "ERROR:\n```text\n" + traceback.format_exc() + "\n```"
|
| 153 |
|
| 154 |
+
# =========================
|
| 155 |
+
# SHAP (safe & bounded)
|
| 156 |
+
# =========================
|
| 157 |
def explain_first_row(file):
|
| 158 |
try:
|
|
|
|
| 159 |
df = _read_table(file)
|
| 160 |
missing, _ = _validate_columns(df)
|
| 161 |
+
if missing:
|
| 162 |
+
return None, f"Missing features for SHAP. First 10: {missing[:10]}"
|
| 163 |
+
|
| 164 |
X = _prepare(df)
|
| 165 |
+
if len(X) == 0:
|
| 166 |
+
return None, "No rows in file."
|
| 167 |
|
| 168 |
+
if IS_PIPELINE:
|
| 169 |
+
X_imp = pd.DataFrame(
|
| 170 |
+
model.named_steps["imputer"].transform(X),
|
| 171 |
+
columns=FEATURES
|
| 172 |
+
)
|
| 173 |
+
clf = model.named_steps["clf"]
|
| 174 |
+
else:
|
| 175 |
+
X_imp = X
|
| 176 |
+
clf = model
|
| 177 |
+
|
| 178 |
+
bg = X_imp.sample(min(50, len(X_imp)), random_state=0)
|
| 179 |
+
|
| 180 |
+
explainer = shap.Explainer(
|
| 181 |
+
clf.predict_proba,
|
| 182 |
+
bg,
|
| 183 |
+
feature_names=FEATURES,
|
| 184 |
+
algorithm="permutation"
|
| 185 |
+
)
|
| 186 |
|
| 187 |
+
explanation = explainer(X_imp.iloc[[0]])
|
| 188 |
+
vals = explanation.values
|
|
|
|
|
|
|
| 189 |
vals_1d = vals[0, :, 1] if vals.ndim == 3 else vals[0]
|
|
|
|
| 190 |
|
| 191 |
+
s = (
|
| 192 |
+
pd.Series(vals_1d, index=FEATURES)
|
| 193 |
+
.sort_values(key=np.abs, ascending=False)
|
| 194 |
+
.head(15)
|
| 195 |
+
)
|
| 196 |
+
|
| 197 |
+
fig = plt.figure(figsize=(7, 5))
|
| 198 |
ax = fig.add_subplot(111)
|
| 199 |
+
s.iloc[::-1].plot(kind="barh", ax=ax)
|
| 200 |
ax.set_title("SHAP values (Sample #1)")
|
| 201 |
fig.tight_layout()
|
| 202 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 203 |
note = (
|
| 204 |
+
"Positive bars increase pathogenic probability; negative bars decrease it.\n\n"
|
| 205 |
+
"SHAP features correspond primarily to ACMG PP3/BP4 (computational evidence).\n"
|
| 206 |
+
"Do not use in isolation for clinical reporting."
|
| 207 |
)
|
| 208 |
+
|
| 209 |
return fig, note
|
| 210 |
+
|
| 211 |
except Exception:
|
| 212 |
+
return None, "ERROR:\n```text\n" + traceback.format_exc() + "\n```"
|
| 213 |
|
| 214 |
+
# =========================
|
| 215 |
+
# UI
|
| 216 |
+
# =========================
|
| 217 |
with gr.Blocks(title="Pathogenicity Predictor") as demo:
|
| 218 |
+
gr.HTML(
|
| 219 |
+
"<h1 style='text-align:center; color:#d63384;'>"
|
| 220 |
+
"Pathogenicity Predictor</h1>"
|
| 221 |
+
)
|
|
|
|
|
|
|
| 222 |
|
| 223 |
+
inp = gr.File(label="Upload CSV / XLSX with trained features")
|
| 224 |
|
|
|
|
|
|
|
| 225 |
with gr.Row():
|
| 226 |
btn_pred = gr.Button("Predict")
|
| 227 |
+
btn_expl = gr.Button("Explain Sample #1 (SHAP)")
|
| 228 |
+
|
| 229 |
out_df = gr.Dataframe(label="Predictions")
|
| 230 |
+
out_plot = gr.Plot(label="SHAP Explanation")
|
| 231 |
out_txt = gr.Markdown()
|
|
|
|
|
|
|
| 232 |
|
| 233 |
+
btn_pred.click(predict_csv, inp, [out_df, out_txt])
|
| 234 |
+
btn_expl.click(explain_first_row, inp, [out_plot, out_txt])
|
| 235 |
+
|
| 236 |
+
# =========================
|
| 237 |
+
# Launch (HF SAFE)
|
| 238 |
+
# =========================
|
| 239 |
+
demo.queue(concurrency_count=1)
|
| 240 |
+
demo.launch()
|