rwe 0.0.1__py3-none-any.whl

rwe/clients/aou.py

@@ -0,0 +1,211 @@
+import os
+import numpy as np
+import pandas as pd
+from tqdm import tqdm
+import multiprocessing as mp
+from scipy import stats
+from docx import Document
+from docx.shared import Inches
+
+import rwe.utils.helpers as uth
+from rwe.plots.variant_info import create_variant_frequency_plots
+from rwe.plots.demographics import demographics_plot
+from rwe.plots.clinical import manhattan
+from rwe.plots.measurements import plot_measurements
+from rwe.plots.surveys import plot_survey_questions
+
+############### Variant Info and Demographics ###############
+def get_individual_info(bucket, chrm, gene, zygosity="hetz"):
+    person_df = pd.read_csv(f"{bucket}/data/rwe_info/raw/person.csv.gz")
+    variant_df = pd.read_csv(f"{bucket}/data/rwe_info/genes/chr{chrm}/{gene}/lof_{zygosity}.csv.gz")
+    person_df["cases"] = person_df.person_id.isin(variant_df[f"{zygosity}_samples"])
+    person_df["group"] = np.where(person_df["cases"], "Cases", "Controls") 
+    return person_df, variant_df
+
+def generate_aou_variant_info_demographics_report(doc: Document, chrm: str, gene: str, zygosity: str ="hetz") -> Document:
+    from rwe.parsers.aou.config import BUCKET
+    person_df, variant_df = get_individual_info(BUCKET, chrm, gene, zygosity)
+    if person_df.empty or variant_df.empty:
+        doc.add_paragraph("No variant or individual level data provided to generate figures.")
+        doc.add_page_break()
+        return doc
+    # 1) Variant information figure
+    fig1, _axes1 = create_variant_frequency_plots(variant_df)
+    fig1_path = uth._save_fig_to_tmp(fig1, basename="variant_information", dpi=300)
+    doc.add_paragraph()  # spacing
+    doc.add_picture(fig1_path, width=Inches(6.5))
+    uth._add_caption(doc, "Figure 1. Variant information summary (carrier counts by consequence; most/least frequent variants).")
+    # 2) Demographics figure
+    fig2, _axes2 = demographics_plot(person_df)
+    fig2_path = uth._save_fig_to_tmp(fig2, basename="demographics", dpi=300)
+    doc.add_paragraph()  # spacing
+    doc.add_picture(fig2_path, width=Inches(6.5))
+    uth._add_caption(doc, "Figure 2. Demographics of cases vs controls (age, ancestry, sex at birth, ethnicity).")
+    return doc
+
+
+############### Clinical Records ###############
+def clean_aou_phewas(phewas_file):
+    df = pd.read_csv(phewas_file, sep="\t")
+    df = df.loc[(df.ancestry == "all")&(df.converged==True)]
+    return df
+
+def get_aou_manhattan(df, gene):
+    name_col = "phecode_string"
+    p_col    = "p_value"
+    odds_ratio_col = "odds_ratio"
+    cat_col  = "phecode_category"
+
+    fig, ax, plot_df = manhattan(
+        df,
+        p_col=p_col,
+        category_col=cat_col,
+        label_col=name_col,
+        odds_ratio_col=odds_ratio_col,
+        sig_p=2.8e-4,
+        top_k_labels=5,
+        title=f"{gene} PheWAS AoU",
+    )
+    return fig, plot_df
+
+def generate_aou_clinical_report(doc, chrm, gene, zygosity):
+    from rwe.parsers.aou.config import BUCKET
+    phewas_file = f"{BUCKET}/data/phewas/results/{chrm}/{gene}_phewas.csv"
+    if uth._gcs_size(phewas_file, BUCKET) > 0: 
+        df = clean_aou_phewas(phewas_file)
+        fig, plot_df = get_aou_manhattan(df, gene)
+        fig_path = uth._save_fig_to_tmp(fig, basename="aou_phewas", dpi=300)
+        doc.add_paragraph()  # spacing
+        doc.add_picture(fig_path, width=Inches(6.5))
+        uth._add_caption(doc, f"Figure X. {gene} PheWAS results from All of Us.")
+        doc.add_paragraph()  # spacing
+
+        # Table 1: Top 10 significant hits (sorted by p-value)
+        top_significant = plot_df.nsmallest(10, 'p_value')
+        columns = ['phecode_string', 'phecode_category', 'odds_ratio', 'p_value']
+        # Rename columns for display
+        display_df = top_significant[columns].copy()
+        display_df.columns = ['Phecode', 'Category', 'OR', 'P-value']
+        uth._add_table_to_doc(doc, display_df, f"Table X. Top 10 significant associations for {gene}",
+                        ['Phecode', 'Category', 'OR', 'P-value'], list(map(Inches, [2.75, 1.75, 0.75, 0.75]))
+                        )
+
+        # Table 2: Top 10 negative beta hits (odds_ratio < 0, sorted by p-value)
+        negative_beta = plot_df[plot_df['odds_ratio'] < 1].nsmallest(10, 'p_value')
+        if len(negative_beta) > 0:
+            display_df_neg = negative_beta[columns].copy()
+            display_df_neg.columns = ['Phecode', 'Category', 'OR', 'P-value']
+            uth._add_table_to_doc(doc, display_df_neg, f"Table X. Top 10 protective associations for {gene}",
+                            ['Phecode', 'Category', 'OR', 'P-value'], list(map(Inches, [2.75, 1.75, 0.75, 0.75]))
+                            )
+
+        doc.add_paragraph()  # spacing
+    else:
+        doc.add_paragraph(f"No AoU PheWAS results found for {gene}.")
+        doc.add_paragraph()  # spacing
+    return doc
+
+
+############### Labs and Measurements ###############
+def remove_outliers_iqr(df, column, multiplier=5):
+    """
+    Remove outliers from a specified column in a pandas DataFrame using the IQR method.
+    """
+    value = np.log1p(df[column].astype(float))
+    Q1 = value.quantile(0.25)
+    Q3 = value.quantile(0.75)
+    IQR = Q3 - Q1
+    lower_bound = Q1 - multiplier * IQR
+    upper_bound = Q3 + multiplier * IQR
+    return np.where(value.between(lower_bound, upper_bound), df[column], pd.NA)
+
+def clean_measurement(df, col="median_value", multiplier=5):
+    df =  df.copy()
+    df = df[df[col] >= 0]
+    df[col] = remove_outliers_iqr(df, col, multiplier)
+    return df.dropna(subset=[col])
+
+def compare(df, measurement):
+    mdf = df.loc[df.measurement==measurement].copy()
+    ctrls = mdf.loc[mdf.cases==False].drop_duplicates(["person_id", "measurement_concept_id"])
+    cases = mdf.loc[mdf.cases==True].drop_duplicates(["person_id", "measurement_concept_id"])
+    ctrls = clean_measurement(ctrls)
+    x = pd.to_numeric(cases["median_value"], errors="coerce").dropna().to_numpy()
+    y = pd.to_numeric(ctrls["median_value"], errors="coerce").dropna().to_numpy()
+    caq1, caq2, caq3 = pd.Series(x).quantile([0.25, 0.5, 0.75])
+    ctq1, ctq2, ctq3 = pd.Series(y).quantile([0.25, 0.5, 0.75])
+    # guard against empty arrays
+    if x.size<20 or y.size<20:
+        return (measurement, np.nan, np.nan, len(cases), len(ctrls), np.nan, np.nan, np.nan)
+    u_stat, p_mwu = stats.mannwhitneyu(x, y, alternative="two-sided")
+    return (measurement, float(u_stat), float(p_mwu), int(len(cases)), int(len(ctrls)), caq1, caq2, caq3, ctq1, ctq2, ctq3)
+
+_G_DF = None
+
+def _init_worker(df):
+    global _G_DF
+    _G_DF = df
+
+def _worker_compare(measurement):
+    # uses global df set once per process
+    return compare(_G_DF, measurement)
+
+def run_parallel(df, measurements=None, n_jobs=None, chunksize=10):
+    if measurements is None:
+        measurements = sorted(df["measurement"].dropna().unique().tolist())
+    if n_jobs is None:
+        n_jobs = max(1, mp.cpu_count() - 1)
+    with mp.Pool(processes=n_jobs, initializer=_init_worker, initargs=(df,)) as pool:
+        rows = list(
+            tqdm(
+                pool.imap_unordered(_worker_compare, measurements, chunksize=chunksize),
+                total=len(measurements),
+                desc="Processing measurements"
+            )
+        )
+    results_df = pd.DataFrame(
+        rows, columns=["measurement", "u_stat", "p_mwu", "n_cases", "n_ctrls", "q1_case", "median_case", "q3_case", "q1_ctrl", "median_ctrl", "q3_ctrl"]
+    )
+    return results_df
+
+def generate_aou_labs_measurements_report(doc: Document, chrm: str, gene: str, zygosity: str) -> Document:
+    from rwe.parsers.aou.config import MEASUREMENT_GROUPS, BUCKET
+    numerical_measurements_df = pd.read_parquet(f"{BUCKET}/data/rwe_info/processed/selected_numerical_measurements.parquet")
+    variant_df = pd.read_csv(f"{BUCKET}/data/rwe_info/genes/chr{chrm}/{gene}/lof_{zygosity}.csv.gz")
+    numerical_measurements_df["cases"] = numerical_measurements_df.person_id.isin(variant_df[f"{zygosity}_samples"])
+    measurements = numerical_measurements_df["measurement"].unique().tolist()
+    res_df = run_parallel(numerical_measurements_df, measurements=measurements, n_jobs=None, chunksize=20)
+    # TODO: add most significant measurements table to doc
+    for k,v in MEASUREMENT_GROUPS.items():
+        f,a,p = plot_measurements(numerical_measurements_df, v, multiplier=5, col="median_value", res_df=res_df)
+        fig_path = uth._save_fig_to_tmp(f, basename=f"aou_measurements_{k}", dpi=300)
+        doc.add_paragraph()  # spacing
+        doc.add_picture(fig_path, width=Inches(6.5))
+        uth._add_caption(doc, f"Figure X. {gene} pLoF carrier {k} results from All of Us.")
+        doc.add_paragraph()  # spacing
+    return doc
+
+############### Surveys ###############
+def clean_aou_surveys(df, survey_col="survey", question_col="question", answer_col="answer_category", zygosity="hetz"):
+    df = df.copy()
+    df[question_col] = df[question_col].str.replace("^" + df[survey_col] + ": ", "", regex=True)
+    df[answer_col] = df[answer_col].str.replace("^" + df[question_col] + ": ", "", regex=True)
+    return df
+
+def generate_aou_survey_report(doc: Document, chrm: str, gene: str, zygosity: str) -> Document:
+    from rwe.parsers.aou.config import BUCKET
+    survey_df = pd.read_parquet(f"{BUCKET}/data/rwe_info/processed/selected_surveys.parquet")
+    variant_df = pd.read_csv(f"{BUCKET}/data/rwe_info/genes/chr{chrm}/{gene}/lof_{zygosity}.csv.gz")
+    survey_df["cases"] = survey_df.person_id.isin(variant_df[f"{zygosity}_samples"])
+    survey_df = clean_aou_surveys(survey_df, zygosity=zygosity)
+    fig, ax = plot_survey_questions(survey_df)
+    fig_path = uth._save_fig_to_tmp(fig, basename="aou_surveys", dpi=300)
+    doc.add_paragraph()  # spacing
+    doc.add_picture(fig_path, width=Inches(6.5))
+    uth._add_caption(doc, f"Figure X. {gene} pLoF carrier survey results from All of Us.")
+    doc.add_paragraph()  # spacing
+    return doc
+
+if __name__ == "__main__":
+    from docx import Document
+    doc = Document()

rwe/clients/azn.py

@@ -0,0 +1,186 @@
+import os
+import shutil
+import zipfile
+from pathlib import Path
+from playwright.sync_api import Playwright, sync_playwright
+from rwe.plots.clinical import manhattan
+import pandas as pd
+
+import rwe.utils.helpers as uth
+from docx.shared import Inches
+
+def extract_only_csv_rename(zip_path: str, gene: str) -> str:
+    """
+    Unzips `zip_path` into its parent directory, keeps only the csv,
+    renames it to {gene}_azphewas.csv, removes everything else including the zip.
+    Returns final csv path.
+    """
+    zip_path = Path(zip_path)
+    gene_dir = zip_path.parent
+    target_csv = gene_dir / f"{gene}_azphewas.csv"
+
+    # Extract zip into gene_dir
+    with zipfile.ZipFile(zip_path, "r") as z:
+        z.extractall(gene_dir)
+
+    # Find CSV(s) extracted
+    csvs = list(gene_dir.rglob("*.csv"))
+    if not csvs:
+        raise FileNotFoundError(f"No CSV found after extracting {zip_path}")
+
+    # If multiple CSVs, pick the largest (usually the main table export)
+    csvs.sort(key=lambda p: p.stat().st_size, reverse=True)
+    chosen = csvs[0]
+
+    # Move/rename chosen CSV to target name at gene_dir root
+    if target_csv.exists():
+        target_csv.unlink()
+    chosen.replace(target_csv)
+
+    # Cleanup: remove everything except the renamed CSV
+    for p in gene_dir.iterdir():
+        if p == target_csv:
+            continue
+        if p.is_dir():
+            shutil.rmtree(p)
+        else:
+            p.unlink()
+
+    return str(target_csv)
+
+
+def run(playwright: Playwright, gene: str, save_path: str) -> None:
+    browser = playwright.chromium.launch(headless=True)
+    context = browser.new_context(accept_downloads=True)
+    page = context.new_page()
+
+    url = f"https://azphewas.com/geneView/6319c068-fd59-46d8-85ee-82d82482eb14/{gene}/glr/binary"
+    page.goto(url, wait_until="networkidle")
+
+    export_control = page.locator('[data-testid="export-settings-control"] button:has-text("Export")')
+    export_control.click()
+
+    inner_export_btn = page.get_by_test_id("export_button")
+    inner_export_btn.wait_for(state="visible")
+
+    plot_checkbox = page.locator('[data-testid="plot-export-type-toggler"] input[type="checkbox"]')
+    plot_checkbox.set_checked(False, force=True)
+    page.wait_for_timeout(300)
+
+    os.makedirs(save_path, exist_ok=True)
+
+    with page.expect_download(timeout=120_000) as dl_info:
+        inner_export_btn.click(force=True)
+
+    download = dl_info.value
+    fname = download.suggested_filename
+    out_dir = save_path
+    os.makedirs(out_dir, exist_ok=True)
+    out_path = os.path.join(out_dir, fname)
+    download.save_as(out_path)
+    print("Saved zip:", out_path)
+
+    final_csv = extract_only_csv_rename(out_path, gene=gene)
+    print("Final CSV:", final_csv)
+
+    context.close()
+    browser.close()
+    return
+
+def clean_azn_phewas(phewas_file):
+    df = pd.read_csv(phewas_file, sep=",")
+    df = df.loc[df["Collapsing model"]=="ptv"]
+    return df
+
+def get_azn_manhattan(gene, phewas_file):
+    df = clean_azn_phewas(phewas_file)
+    name_col = "Phenotype"
+    p_col    = "P value"
+    odds_ratio_col = "Odds ratio"
+    cat_col  = "Phenotypic category"
+    AZN_CHAPTER_SHORT = {
+        "Chapter I Certain infectious and parasitic diseases": "Infectious",
+        "Chapter II Neoplasms": "Neoplasms",
+        "Chapter III Diseases of the blood and blood-forming organs and certain disorders involving the immune mechanism": "Blood/Immune",
+        "Chapter IV Endocrine nutritional and metabolic diseases": "Endocrine/Metabolic",
+        "Chapter V Mental and behavioural disorders": "Mental/Behavioral",
+        "Chapter VI Diseases of the nervous system": "Nervous System",
+        "Chapter VII Diseases of the eye and adnexa": "Eye",
+        "Chapter VIII Diseases of the ear and mastoid process": "Ear/Mastoid",
+        "Chapter IX Diseases of the circulatory system": "Circulatory",
+        "Chapter X Diseases of the respiratory system": "Respiratory",
+        "Chapter XI Diseases of the digestive system": "Digestive",
+        "Chapter XII Diseases of the skin and subcutaneous tissue": "Skin",
+        "Chapter XIII Diseases of the musculoskeletal system and connective tissue": "Musculoskeletal",
+        "Chapter XIV Diseases of the genitourinary system": "Genitourinary",
+        "Chapter XV Pregnancy childbirth and the puerperium": "Pregnancy/Childbirth",
+        "Chapter XVII Congenital malformations deformations and chromosomal abnormalities": "Congenital",
+        "Chapter XVIII Symptoms signs and abnormal clinical and laboratory findings not elsewhere classified": "Symptoms/Findings",
+        "Chapter XXI Factors influencing health status and contact with health services": "Health Services",
+    }
+
+    df[name_col] = df[name_col].astype(str).str.rsplit("#", n=1).str[-1]
+    df[cat_col] = df[cat_col].map(AZN_CHAPTER_SHORT).fillna("Other")
+
+    fig, ax, plot_df = manhattan(
+        df,
+        p_col=p_col,
+        category_col=cat_col,
+        label_col=name_col,
+        odds_ratio_col=odds_ratio_col,
+        sig_p=2.8e-4,
+        top_k_labels=5,
+        title=f"{gene} PheWAS AstraZeneca",
+    )
+    return fig, plot_df
+
+def generate_azn_clinical_report(doc, gene, phewas_dir, phewas_filename=""):
+    if not phewas_filename:
+        phewas_filename = f"{gene}_azphewas.csv"
+    phewas_file = os.path.join(phewas_dir, phewas_filename)
+
+    if not os.path.exists(phewas_file):
+        with sync_playwright() as playwright:
+            run(playwright, gene=gene, save_path=phewas_dir)
+
+    if os.path.exists(phewas_file): 
+        fig, plot_df = get_azn_manhattan(gene, phewas_file)
+        fig_path = uth._save_fig_to_tmp(fig, basename="azn_phewas", dpi=300)
+        doc.add_paragraph()  # spacing
+        doc.add_picture(fig_path, width=Inches(6.5))
+        uth._add_caption(doc, f"Figure X. {gene} PheWAS results from AstraZeneca.")
+        doc.add_paragraph()  # spacing
+
+        # Table 1: Top 10 significant hits (sorted by p-value)
+        top_significant = plot_df.nsmallest(10, 'P value')
+        columns = ['Phenotype', 'Phenotypic category', 'Odds ratio', 'P value']
+        # Rename columns for display
+        display_df = top_significant[columns].copy()
+        display_df.columns = ['Phenotype', 'Category', 'OR', 'P-value']
+        uth._add_table_to_doc(doc, display_df, f"Table X. Top 10 significant associations for {gene}",
+                        ['Phenotype', 'Category', 'OR', 'P-value'], list(map(Inches, [2.75, 1.75, 0.75, 0.75]))
+                        )
+        # Table 2: Top 10 negative beta hits (odds_ratio < 0, sorted by p-value)
+        negative_beta = plot_df[plot_df['Odds ratio'] < 1].nsmallest(10, 'P value')
+        if len(negative_beta) > 0:
+            display_df_neg = negative_beta[columns].copy()
+            display_df_neg.columns = ['Phenotype', 'Category', 'OR', 'P-value']
+            uth._add_table_to_doc(doc, display_df_neg, f"Table X. Top 10 protective associations for {gene}",
+                            ['Phenotype', 'Category', 'OR', 'P-value'], list(map(Inches, [2.75, 1.75, 0.75, 0.75]))
+                            )
+        doc.add_paragraph()  # spacing
+
+    else:
+        doc.add_paragraph(f"No AstraZeneca PheWAS results found for {gene}.")
+        doc.add_paragraph()  # spacing
+    return doc
+
+
+if __name__ == "__main__":
+    from docx import Document
+    doc = Document()
+    gene = "INHBE"
+    phewas_dir = "/home/dbanerjee/deepro/rwe/data/test/"
+    doc = generate_azn_clinical_report(doc, gene, phewas_dir)
+    doc.save("/home/dbanerjee/deepro/rwe/data/test/INHBE_azn_report.docx")
+

rwe/clients/genebass.py

@@ -0,0 +1,137 @@
+import os
+import numpy as np
+from playwright.sync_api import Playwright, sync_playwright
+import random
+import pandas as pd
+from docx.shared import Inches, Pt
+
+from rwe.plots.clinical import manhattan
+import rwe.utils.helpers as uth
+from rwe.clients.hgnc import get_gene_info
+
+def run(playwright: Playwright, gene:str, gene_id:str, save_path:str) -> None:
+    browser = playwright.chromium.launch(headless=True)
+    context = browser.new_context(accept_downloads=True)
+    # Open new page
+    page = context.new_page()
+    # Go to https://app.genebass.org/
+   # page.goto("https://app.genebass.org/")
+    query_id = "https://app.genebass.org/gene/" + gene_id +"?burdenSet=pLoF&phewasOpts=1&resultLayout=full"
+    page.goto(query_id)
+    page.click("label:has-text(\"Burden\")")
+    page.wait_for_timeout(random.uniform(100, 1000))
+    # Click text=Export data to CSV
+    with page.expect_download() as download_info:
+        page.click("text=Export data to CSV")
+    download = download_info.value
+    download_path = os.path.join(save_path, f"{gene}_genebass.csv")
+    os.makedirs(os.path.dirname(download_path), exist_ok=True)
+    download.save_as(path=download_path)
+    page.wait_for_timeout(random.uniform(100, 1000))
+    # ---------------------
+    context.close()
+    browser.close()
+    return
+
+def clean_genebass_phewas(phewas_file):
+    df = pd.read_csv(phewas_file, sep=",")
+    df = df.loc[df["Trait type"]=="icd_first_occurrence"]
+    return df
+
+def get_genebass_manhattan(gene, phewas_file):
+    df = clean_genebass_phewas(phewas_file)
+    name_col = "Description"
+    p_col    = "P-Value (Burden)"
+    odds_ratio_col = None
+    beta_col = "Beta"
+    cat_col  = "Category"
+
+    df[p_col] = pd.to_numeric(df[p_col], errors="coerce")
+    df = df[df[p_col].notna() & (df[p_col] > 0) & np.isfinite(df[p_col])]
+    prefix = "Health-related outcomes > First occurrences > "
+    df[cat_col] = (df[cat_col].astype(str)
+                    .str.replace(prefix, "", regex=False)
+                    .fillna("Unknown"))
+    GENEBASS_CATEGORY_SHORT = {
+        "Blood, blood-forming organs and certain immune disorders": "Blood/Immune",
+        "Certain conditions originating in the perinatal period": "Perinatal",
+        "Certain infectious and parasitic diseases": "Infectious",
+        "Circulatory system disorders": "Circulatory",
+        "Congenital disruptions and chromosomal abnormalities": "Congenital",
+        "Digestive system disorders": "Digestive",
+        "Ear and mastoid process disorders": "Ear/Mastoid",
+        "Endocrine, nutritional and metabolic diseases": "Endocrine/Metabolic",
+        "Eye and adnexa disorders": "Eye",
+        "Genitourinary system disorders": "Genitourinary",
+        "Mental and behavioural disorders": "Mental/Behavioral",
+        "Musculoskeletal system and connective tissue disorders": "Musculoskeletal",
+        "Nervous system disorders": "Nervous System",
+        "Pregnancy, childbirth and the puerperium": "Pregnancy/Childbirth",
+        "Respiratory system disorders": "Respiratory",
+        "Skin and subcutaneous tissue disorders": "Skin",
+    }
+    df[cat_col] = df[cat_col].map(GENEBASS_CATEGORY_SHORT).fillna(df[cat_col])
+    fig, ax, plot_df = manhattan(
+        df,
+        p_col=p_col,
+        category_col=cat_col,
+        label_col=name_col,
+        beta_col=beta_col,
+        sig_p=2.8e-4,
+        top_k_labels=5,
+        title=f"{gene} PheWAS GeneBass",
+    )
+    return fig, plot_df
+
+def generate_genebass_clinical_report(doc, gene, phewas_dir, phewas_filename=""):
+    if not phewas_filename:
+        phewas_filename = f"{gene}_genebass.csv"
+    phewas_file = os.path.join(phewas_dir, phewas_filename)
+
+    if not os.path.exists(phewas_file):
+        hgnc_path = os.path.join(phewas_dir, "gene_with_protein_product.txt")
+        approved_symbol, ensembl_id, entrez_id = get_gene_info(gene, hgnc_path)
+        with sync_playwright() as playwright:
+            run(playwright, approved_symbol, ensembl_id, save_path=phewas_dir)
+
+    if os.path.exists(phewas_file):
+        fig, plot_df = get_genebass_manhattan(gene, phewas_file)
+        fig_path = uth._save_fig_to_tmp(fig, basename="genebass_phewas", dpi=300)
+        doc.add_paragraph()  # spacing
+        doc.add_picture(fig_path, width=Inches(6))
+        uth._add_caption(doc, f"Figure X. {gene} PheWAS results from GeneBass.")
+        doc.add_paragraph()  # spacing
+
+        # Table 1: Top 10 significant hits (sorted by p-value)
+        top_significant = plot_df.nsmallest(10, 'P-Value (Burden)')
+        columns = ['Description', 'Category', 'Beta', 'P-Value (Burden)']
+        # Rename columns for display
+        display_df = top_significant[columns].copy()
+        display_df.columns = ['Description', 'Category', 'Beta', 'P-value']
+        uth._add_table_to_doc(doc, display_df, f"Table X. Top 10 significant associations for {gene}",
+                        ['Description', 'Category', 'Beta', 'P-value'], list(map(Inches, [2.75, 2.0, 0.5, 0.75]))
+                        )
+
+        # Table 2: Top 10 negative beta hits (beta < 0, sorted by p-value)
+        negative_beta = plot_df[plot_df['Beta'] < 0].nsmallest(10, 'P-Value (Burden)')
+        if len(negative_beta) > 0:
+            display_df_neg = negative_beta[columns].copy()
+            display_df_neg.columns = ['Description', 'Category', 'Beta', 'P-value']
+            uth._add_table_to_doc(doc, display_df_neg, f"Table X. Top 10 protective associations for {gene}",
+                            ['Description', 'Category', 'Beta', 'P-value'], list(map(Inches, [2.75, 2.0, 0.5, 0.75]))
+                            )
+            
+        doc.add_paragraph()  # spacing
+    else:
+        doc.add_paragraph(f"No GeneBass PheWAS results found for {gene}.")
+        doc.add_paragraph()  # spacing
+    return doc
+
+if __name__ == "__main__":
+    from docx import Document
+    doc = Document()
+    gene = "INHBE"
+    phewas_dir = "/home/dbanerjee/deepro/rwe/data/test/"
+    doc = generate_genebass_clinical_report(doc, gene, phewas_dir)
+    doc.save("/home/dbanerjee/deepro/rwe/data/test/INHBE_genebass_report.docx")
+

rwe/clients/hgnc.py

@@ -0,0 +1,52 @@
+import pandas as pd
+import os
+import requests
+
+def load_data(hgnc_path):
+    # Load HGNC gene list (protein-coding genes)
+    if os.path.exists(hgnc_path):
+        hgnc_df = pd.read_csv(hgnc_path, sep='\t', dtype=str)
+    else:
+        url = 'https://storage.googleapis.com/public-download-files/hgnc/tsv/tsv/locus_types/gene_with_protein_product.txt'
+        resp = requests.get(url)
+        resp.raise_for_status()
+        with open(hgnc_path, 'wb') as f:
+            f.write(resp.content)
+        hgnc_df = pd.read_csv(hgnc_path, sep='\t', dtype=str)
+
+    return hgnc_df
+
+
+def build_gene_symbol_map(hgnc_df):
+    """
+    Map all approved, alias, and previous symbols to the current HGNC symbol.
+    """
+    symbol_map = {}
+    for _, row in hgnc_df.iterrows():
+        approved = row.get('symbol')
+        if pd.isna(approved):
+            continue
+        symbol_map[approved] = approved
+        for col in ['alias_symbol', 'prev_symbol']:
+            if col in row and pd.notna(row[col]):
+                for alias in row[col].split('|'):
+                    symbol_map[alias] = approved
+    ensembl_map = dict(zip(hgnc_df['symbol'], hgnc_df['ensembl_gene_id']))
+    entrez_map = dict(zip(hgnc_df['symbol'], hgnc_df['entrez_id']))
+    return symbol_map, ensembl_map, entrez_map
+
+def get_gene_info(gene, hgnc_path):
+    hgnc_df = load_data(hgnc_path)
+    symbol_map, ensembl_map, entrez_map = build_gene_symbol_map(hgnc_df)
+    approved_symbol = symbol_map.get(gene)
+    if not approved_symbol:
+        raise ValueError(f"Gene symbol '{gene}' not found in HGNC data.")
+    ensembl_id = ensembl_map.get(approved_symbol)
+    entrez_id = entrez_map.get(approved_symbol)
+    return approved_symbol, ensembl_id, entrez_id
+
+if __name__ == "__main__":
+    hgnc_path = "/home/dbanerjee/deepro/rwe/data/hgnc/gene_with_protein_product.txt"
+    approved_symbol, ensembl_id, entrez_id = get_gene_info("ANGPTL3", hgnc_path)
+    print(f"Approved symbol: {approved_symbol}, Ensembl ID: {ensembl_id}, Entrez ID: {entrez_id}")
+

rwe/generate_report.py

@@ -0,0 +1,66 @@
+import os
+import pandas as pd
+
+import matplotlib
+matplotlib.use("Agg")  # important for batch/report generation (no display)
+
+import rwe.utils.report as utr
+
+
+
+# ---------- Main user-facing function ----------
+
+def generate_rwe_report(
+    gene: str,
+    chrm: str,
+    zygosity: str = "hetz",
+    out_docx_path: str = "",
+    logo_path: str = "",
+    report_date: str = "",
+):
+    """
+    Creates a DOCX report with:
+      - Title page
+      - Contents (TOC field)
+      - Five section headers
+      - Variant information + Demographics figures inserted into section 1
+
+    Parameters
+    ----------
+    gene : str
+    chrm : str
+    out_docx_path : str
+    logo_path : optional str (path to Arrowhead logo PNG/JPG)
+    report_date : optional str (e.g. "January 26th, 2026")
+    """
+    proj_dir = os.path.dirname(out_docx_path) or "."
+    os.makedirs(proj_dir, exist_ok=True)
+
+    # --- Title and Contents --- #
+    doc = utr.generate_title_and_contents(gene, logo_path=logo_path, report_date=report_date)
+
+    # --- Variant information and demographics --- #
+    doc = utr.generate_variant_information_and_demographics(doc, chrm, gene, zygosity)
+
+    # --- Clinical records --- #
+    doc = utr.generate_clinical_records(doc, chrm, gene, zygosity, proj_dir)
+    
+    # --- Labs and measurements --- #
+    doc = utr.generate_labs_and_measurements(doc, chrm, gene, zygosity, proj_dir)
+
+    # --- Survey information --- #
+    doc = utr.generate_survey_information(doc, chrm, gene, zygosity, proj_dir)
+
+    # --- Homozygous loss of function carriers --- #
+    doc = utr.generate_homozygous_lof_carriers(doc, chrm, gene, zygosity, proj_dir)
+
+
+    # Save
+    if not out_docx_path:
+        out_docx_path = f"{gene}_RWE_report.docx"
+    doc.save(out_docx_path)
+    return out_docx_path
+
+if __name__ == "__main__":
+    import pandas as pd
+