app.py

import gradio as gr
import pandas as pd
import requests
import rpy2.robjects as robjects


def get_SNP():
    r_code = """
        library(GWAScFDR)
        library(dplyr)
        library(data.table)

        rm(list = ls())

        AD <- fread("Lambert/IGAP_summary_statistics/IGAP_stage_1.txt", header = TRUE, stringsAsFactors = F)
        AMD_F <- fread("Fritsche/Fritsche-26691988.txt.gz", header = TRUE, stringsAsFactors = F)

        AD <- AD[as.numeric(AD$Pvalue) < 0.00001, ]
        AMD_F <- AMD_F[as.numeric(AMD_F$GC.Pvalue) < 0.00001, ]

        names(AD)[c(3, 8)] <- c("Marker", "p_AD")
        names(AMD_F)[c(1, 8)] <- c("Marker", "p_AMDF")

        # merge data
        AD_AMDF <- merge(AD, AMD_F, by = "Marker", all = FALSE)

        # calculate cFDR
        AD_AMDF$FDR_AD_AMD <- cFDR(as.numeric(AD_AMDF$p_AD), as.numeric(AD_AMDF$p_AMDF))
        AD_AMDF$FDR_AMD_AD <- cFDR(as.numeric(AD_AMDF$p_AMDF), as.numeric(AD_AMDF$p_AD))

        # read package to calculate ccFDR (the bigher one)
        AD_AMDF$ccFDR <- pmax(AD_AMDF$FDR_AD_AMD, AD_AMDF$FDR_AMD_AD)

        AD_AMDF <- AD_AMDF[AD_AMDF$ccFDR < 0.001, ]

        # write.csv(AD_AMDF, file = "ccFDR_AD_AMDF.csv")
        
        AD_AMDF_rsid <- AD_AMDF$Marker
        """

    robjects.r(r_code)
    rsidlist = list(robjects.globalenv["AD_AMDF_rsid"])
    
    return rsidlist


def get_SNP_gene(rsidlist):
    rsid_gene = {}
    for rsid in rsidlist:
        url = f"https://www.ebi.ac.uk/gwas/rest/api/singleNucleotidePolymorphisms/{rsid}"
        print(url)
        try:
            response = requests.get(url)
            response.raise_for_status()
            data = response.json()

            genelist = []
            for context in data["genomicContexts"]:
                if context["distance"] != 0:
                    continue
                else:
                    genelist.append(context["gene"]["geneName"])
            genelist = list(set(genelist))

            rsid_gene[rsid] = genelist
        except requests.exceptions.RequestException as e:
            print(f"Error occurred while fetching data for {rsid}: {str(e)}")
            continue

    return rsid_gene

def todf(disease1, disease2):
    if disease1 != "" and disease2 != "":
        rsidlist = get_SNP()
        rsid_gene = get_SNP_gene(rsidlist)

        for key, value in rsid_gene.items():
            if value == []:
                rsid_gene[key] = ["None"]
        df = pd.DataFrame(list(rsid_gene.items()), columns=['rsid', 'genes'])
        df['genes'] = df['genes'].apply(lambda x: x[0] if isinstance(x, list) and len(x) > 0 else x)

    return df

def visible(identifiers):
    identifiers = identifiers.split("\n")
    identifiers = list(filter(None, identifiers))
    if len(identifiers) > 0:
        return topo_para.update(visible=True), network_stats.update(
            visible=True), enrichment.update(
                visible=True), button_download.update(visible=True)


# 使用gr.Blocks()创建和组合组件
with gr.Blocks() as demo:
    with gr.Row():
        with gr.Column(scale=1):
            # 创建输入组件
            disease1 = gr.Dropdown(choices=['Alzheimer disease'], value='Alzheimer disease', label='Disease 1')
            disease2 = gr.Dropdown(choices=['age-related macular degeneration'], value='age-related macular degeneration', label='Disease 2')

            # 创建按钮组件
            with gr.Row():
                # button_clear = gr.Button("Clear", elem_id="clear-button")
                button_input = gr.Button("Submit", elem_id="submit-button")

            # 临时组件
            cache = gr.Textbox(visible=False)

        with gr.Column(scale=3.25, min_width=985):
            rsid_gene_df = gr.Dataframe(label="Variants and Genes:", elem_id="topo-para", visible=True)

    # 按钮监听
    button_input.click(fn=todf, inputs=[disease1, disease2], outputs=rsid_gene_df)

# 在本地运行demo对象
demo.launch(share=False)