app.py

import gradio as gr

from concurrency import execute_multithread
from get_index import get_engines
from protein_viz import get_gene_name, get_protein_name, render_html

index_repo = "ronig/siamese_protein_index"
model_repo = "ronig/protein_search_engine"
engines = get_engines(index_repo, model_repo)
available_indexes = list(engines.keys())
app_description = """
# Protein Binding Search Engine
This application enables a quick protein-peptide binding search based on sequences. 
You can use it to search the full [PDB](https://www.rcsb.org/) database or in a specific organism genome.
"""


def search_and_display(seq, n_res, index_selection):
    n_res = int(limit_n_results(n_res))
    engine = engines[index_selection]
    search_res = engine.search_by_sequence(seq, n=n_res)
    results_options = update_dropdown_menu(search_res)
    formatted_search_results = format_search_results(search_res)
    return formatted_search_results, results_options


def limit_n_results(n):
    return max(min(n, 20), 1)


def update_dropdown_menu(search_res):
    choices = []
    for row in search_res:
        if "pdb_name" in row and "chain_id" in row:
            choice = ".".join([row["pdb_name"], row["chain_id"]])
            choices.append(choice)
    if choices:
        update = gr.Dropdown.update(
            choices=choices, interactive=True, value=choices[0], visible=True
        )
    else:
        update = gr.Dropdown.update(
            choices=choices, interactive=True, visible=False, value=None
        )
    return update


def format_search_results(raw_search_results):
    formatted_search_results = {}
    for key, value in execute_multithread(
        func=format_search_result,
        inputs=({"raw_result": res} for res in raw_search_results),
        n_workers=len(raw_search_results),
    ):
        formatted_search_results[key] = value
    return formatted_search_results


def format_search_result(raw_result):
    is_pdb = "pdb_name" in raw_result
    if is_pdb:
        key, value = parse_pdb_search_result(raw_result)
    else:
        key, value = parse_fasta_search_result(raw_result)
    return key, value


def parse_fasta_search_result(raw_result):
    gene = parse_gene_from_fasta_entry(raw_result["description"])
    key = f"Gene: {gene}"
    value = raw_result["score"]
    return key, value


def parse_pdb_search_result(raw_result):
    prot = raw_result["pdb_name"]
    chain = raw_result["chain_id"]
    value = raw_result["score"]
    gene_name, species = get_gene_name(pdb_id=prot, chain_id=chain)
    key = f"PDB: {prot}.{chain}"
    if gene_name is not None:
        key += f" | Gene: {gene_name} | Organism: {species}"
    return key, value


def parse_gene_from_fasta_entry(description):
    after = description.split("GN=")[1]
    gene = after.split(" ")[0]
    return gene


def switch_viz(new_choice):
    if new_choice is None:
        html = ""
        title_update = gr.Markdown.update(visible=False)
        description_update = gr.Markdown.update(value=None, visible=False)
    else:
        choice_parts = new_choice.split(".")
        pdb_id, chain = choice_parts[0], choice_parts[1]
        title_update = gr.Markdown.update(visible=True)
        protein_name = get_protein_name(pdb_id)

        new_value = f"""**PDB Title**: {protein_name}"""

        description_update = gr.Markdown.update(value=new_value, visible=True)
        html = render_html(pdb_id=pdb_id, chain=chain)
    return html, title_update, description_update


with gr.Blocks() as demo:
    with gr.Column():
        gr.Markdown(app_description)
        with gr.Column():
            with gr.Row():
                with gr.Column():
                    seq_input = gr.Textbox(value="APTMPPPLPP", label="Input Sequence")
                    n_results = gr.Number(5, label="N Results")
                    index_selector = gr.Dropdown(
                        choices=available_indexes,
                        value="Pdb",
                        multiselect=False,
                        visible=True,
                        label="Index",
                    )
                    search_button = gr.Button("Search", variant="primary")
                search_results = gr.Label(num_top_classes=20, label="Search Results")
            viz_header = gr.Markdown("## Visualization", visible=False)
            results_selector = gr.Dropdown(
                choices=[],
                multiselect=False,
                visible=False,
                label="Visualized Search Result",
            )
            viz_body = gr.Markdown("", visible=False)
            protein_viz = gr.HTML(
                value=render_html(pdb_id=None, chain=None),
                label="Protein Visualization",
            )
            gr.Examples(
                ["APTMPPPLPP", "KFLIYQMECSTMIFGL", "PHFAMPPIHEDHLE", "AEERIISLD"],
                inputs=[seq_input],
            )
    search_button.click(
        search_and_display,
        inputs=[seq_input, n_results, index_selector],
        outputs=[search_results, results_selector],
    )
    results_selector.change(
        switch_viz, inputs=results_selector, outputs=[protein_viz, viz_header, viz_body]
    )

if __name__ == "__main__":
    demo.launch()