utils.py

# %%
import os
from typing import Any, Dict, List

import pandas as pd
import requests
import streamlit as st
from sentence_transformers import SentenceTransformer
from sqlalchemy import create_engine, text

username = "demo"
password = "demo"
hostname = os.getenv("IRIS_HOSTNAME", "localhost")
port = "1972"
namespace = "USER"
CONNECTION_STRING = f"iris://{username}:{password}@{hostname}:{port}/{namespace}"

engine = create_engine(CONNECTION_STRING)


def get_all_diseases_name(engine) -> List[List[str]]:
    print("Fetching all disease names...")
    with engine.connect() as conn:
        with conn.begin():
            sql = f"""
                    SELECT label FROM Test.EntityEmbeddings
                    """
            result = conn.execute(text(sql))
            data = result.fetchall()

    all_diseases = [row[0] for row in data if row[0] != "nan"]
    return all_diseases


def get_uri_from_name(engine, name: str) -> str:
    with engine.connect() as conn:
        with conn.begin():
            sql = f"""
                    SELECT uri FROM Test.EntityEmbeddings
                    WHERE label = '{name}'
                    """
            result = conn.execute(text(sql))
            data = result.fetchall()
    return data[0][0].split("/")[-1]


def get_most_similar_diseases_from_uri(
    engine, original_disease_uri: str, threshold: float = 0.8
) -> List[str]:
    with engine.connect() as conn:
        with conn.begin():
            sql = f"""
                    SELECT * FROM Test.EntityEmbeddings
                    """
            result = conn.execute(text(sql))
            data = result.fetchall()

    all_diseases = [row[1] for row in data if row[1] != "nan"]
    return all_diseases


def get_uri_from_name(engine, name: str) -> str:
    with engine.connect() as conn:
        with conn.begin():
            sql = f"""
                    SELECT uri FROM Test.EntityEmbeddings
                    WHERE label = '{name}'
                    """
            result = conn.execute(text(sql))
            data = result.fetchall()
    return data[0][0].split("/")[-1]


def get_most_similar_diseases_from_uri(
    engine, original_disease_uri: str, threshold: float = 0.8
) -> List[str]:
    with engine.connect() as conn:
        with conn.begin():
            sql = f"""
                    SELECT TOP 10 e1.uri AS uri1, e2.uri AS uri2, e1.label AS label1, e2.label AS label2,
                    VECTOR_COSINE(e1.embedding, e2.embedding) AS distance
                    FROM Test.EntityEmbeddings e1, Test.EntityEmbeddings e2
                    WHERE e1.uri = 'http://identifiers.org/medgen/{original_disease_uri}'
                    AND VECTOR_COSINE(e1.embedding, e2.embedding) > {threshold}
                    AND e1.uri != e2.uri
                    ORDER BY distance DESC
                    """
            result = conn.execute(text(sql))
            data = result.fetchall()

    similar_diseases = [
        (row[1].split("/")[-1], row[3], row[4]) for row in data if row[3] != "nan"
    ]
    return similar_diseases


def get_clinical_record_info(clinical_record_id: str) -> Dict[str, Any]:
    # Request:
    # curl -X GET "https://clinicaltrials.gov/api/v2/studies/NCT00841061" \
    # -H "accept: text/csv"
    request_url = f"https://clinicaltrials.gov/api/v2/studies/{clinical_record_id}"
    response = requests.get(request_url, headers={"accept": "application/json"})
    return response.json()


def get_clinical_records_by_ids(clinical_record_ids: List[str]) -> List[Dict[str, Any]]:
    clinical_records = []
    for clinical_record_id in clinical_record_ids:
        clinical_record_info = get_clinical_record_info(clinical_record_id)
        clinical_records.append(clinical_record_info)
    return clinical_records


def get_similarities_among_diseases_uris(
    uri_list: List[str],
) -> List[tuple[str, str, float]]:
    uri_list = ", ".join([f"'{uri}'" for uri in uri_list])
    with engine.connect() as conn:
        with conn.begin():
            sql = f"""
                    SELECT e1.uri AS uri1, e2.uri AS uri2, VECTOR_COSINE(e1.embedding, e2.embedding) AS distance
                    FROM Test.EntityEmbeddings e1, Test.EntityEmbeddings e2
                    WHERE e1.uri IN ({uri_list}) AND e2.uri IN ({uri_list}) AND e1.uri != e2.uri
                """
            result = conn.execute(text(sql))
            data = result.fetchall()

    return [
        {
            "uri1": row[0].split("/")[-1],
            "uri2": row[1].split("/")[-1],
            "distance": float(row[2]),
        }
        for row in data
    ]


def augment_the_set_of_diseaces(diseases: List[str]) -> str:
    augmented_diseases = diseases.copy()
    for i in range(10 - len(augmented_diseases)):
        with engine.connect() as conn:
            with conn.begin():
                sql = f"""
                    SELECT TOP 1 e2.uri AS new_disease, (SUM(VECTOR_COSINE(e1.embedding, e2.embedding))/ {len(augmented_diseases)})  AS score
                    FROM Test.EntityEmbeddings e1, Test.EntityEmbeddings e2
                    WHERE e1.uri IN ({','.join([f"'{disease}'" for disease in augmented_diseases])})
                    AND e2.uri NOT IN ({','.join([f"'{disease}'" for disease in augmented_diseases])})
                    AND e2.label != 'nan'
                    GROUP BY e2.label
                    ORDER BY score DESC
                    """

                result = conn.execute(text(sql))
                data = result.fetchall()

        augmented_diseases.append(data[0][0])

    return augmented_diseases


def get_embedding(string: str, encoder) -> List[float]:
    # Embed the string using sentence-transformers
    vector = encoder.encode(string, show_progress_bar=False)
    return vector


def get_diseases_related_to_a_textual_description(
    description: str, encoder
) -> List[str]:
    # Embed the description using sentence-transformers
    description_embedding = get_embedding(description, encoder)
    string_representation = str(description_embedding.tolist())[1:-1]

    with engine.connect() as conn:
        with conn.begin():
            sql = f"""
                    SELECT TOP 10 d.uri, VECTOR_COSINE(d.embedding, TO_VECTOR('{string_representation}', DOUBLE)) AS distance
                    FROM Test.DiseaseDescriptions d
                    ORDER BY distance DESC
                """
            result = conn.execute(text(sql))
            data = result.fetchall()

    return [
        {"uri": row[0], "distance": float(row[1])}
        for row in data
        if float(row[1]) > 0.8
    ]


def get_clinical_trials_related_to_diseases(diseases: List[str], encoder) -> List[str]:
    # Embed the diseases using sentence-transformers
    diseases_string = ", ".join(diseases)
    disease_embedding = get_embedding(diseases_string, encoder)
    string_representation = str(disease_embedding.tolist())[1:-1]

    with engine.connect() as conn:
        with conn.begin():
            sql = f"""
                    SELECT TOP 20 d.nct_id, VECTOR_COSINE(d.embedding, TO_VECTOR('{string_representation}', DOUBLE)) AS distance
                    FROM Test.ClinicalTrials d
                    ORDER BY distance DESC
                """
            result = conn.execute(text(sql))
            data = result.fetchall()

    return [{"nct_id": row[0], "distance": row[1]} for row in data]


def get_similarities_df(diseases: List[Dict[str, Any]]) -> pd.DataFrame:
    # Find out the score of each disease by averaging the cosine similarity of the embeddings of the diseases that include it as uri1 or uri2
    df_diseases_similarities = pd.DataFrame(diseases)
    # Use uri1 as the index, and uri2 as the columns. The values are the distances.
    df_diseases_similarities = df_diseases_similarities.pivot(
        index="uri1", columns="uri2", values="distance"
    )
    # Fill the diagonal with 1.0
    df_diseases_similarities = df_diseases_similarities.fillna(1.0)

    return df_diseases_similarities


def filter_out_less_promising_diseases(info_dicts: List[Dict[str, Any]]) -> List[str]:
    df_diseases_similarities = get_similarities_df(info_dicts)

    # Filter out the diseases that are 0.2 standard deviations below the mean
    mean = df_diseases_similarities.mean().mean()
    std = df_diseases_similarities.mean().std()
    filtered_diseases = df_diseases_similarities.mean()[
        df_diseases_similarities.mean() > mean - 0.2 * std
    ].index.tolist()
    return filtered_diseases, df_diseases_similarities


def get_labels_of_diseases_from_uris(uris: List[str]) -> List[str]:
    with engine.connect() as conn:
        with conn.begin():
            joined_uris = ", ".join([f"'{uri}'" for uri in uris])
            sql = f"""
                    SELECT label FROM Test.EntityEmbeddings
                    WHERE uri IN ({joined_uris})
                """
            result = conn.execute(text(sql))
            data = result.fetchall()

    return [row[0] for row in data]


def to_capitalized_case(string: str) -> str:
    string = string.replace("_", " ")
    if string.isupper():
        return string[0] + string[1:].lower()


def list_to_capitalized_case(strings: List[str]) -> str:
    strings = [to_capitalized_case(s) for s in strings]
    return ", ".join(strings)


def render_trial_details(trial: dict) -> None:
    # TODO: handle key errors for all cases (→ do not render)

    official_title = trial["protocolSection"]["identificationModule"]["officialTitle"]
    st.write(f"##### {official_title}")

    try:
        st.write(trial["protocolSection"]["descriptionModule"]["briefSummary"])
    except KeyError:
        try:
            st.write(
                trial["protocolSection"]["descriptionModule"]["detailedDescription"]
            )
        except KeyError:
            st.error("No description available.")

    st.write("###### Status")
    try:
        status_module = {
            "Status": to_capitalized_case(
                trial["protocolSection"]["statusModule"]["overallStatus"]
            ),
            "Status Date": trial["protocolSection"]["statusModule"][
                "statusVerifiedDate"
            ],
            "Has Results": trial["hasResults"],
        }
        st.table(status_module)
    except KeyError:
        st.info("No status information available.")

    st.write("###### Design")
    try:
        design_module = {
            "Study Type": to_capitalized_case(
                trial["protocolSection"]["designModule"]["studyType"]
            ),
            "Phases": list_to_capitalized_case(
                trial["protocolSection"]["designModule"]["phases"]
            ),
            "Allocation": to_capitalized_case(
                trial["protocolSection"]["designModule"]["designInfo"]["allocation"]
            ),
            "Primary Purpose": to_capitalized_case(
                trial["protocolSection"]["designModule"]["designInfo"]["primaryPurpose"]
            ),
            "Participants": trial["protocolSection"]["designModule"]["enrollmentInfo"][
                "count"
            ],
            "Masking": to_capitalized_case(
                trial["protocolSection"]["designModule"]["designInfo"]["maskingInfo"][
                    "masking"
                ]
            ),
            "Who Masked": list_to_capitalized_case(
                trial["protocolSection"]["designModule"]["designInfo"]["maskingInfo"][
                    "whoMasked"
                ]
            ),
        }
        st.table(design_module)
    except KeyError:
        st.info("No design information available.")

    st.write("###### Interventions")
    try:
        interventions_module = {}
        for intervention in trial["protocolSection"]["armsInterventionsModule"][
            "interventions"
        ]:
            name = intervention["name"]
            desc = intervention["description"]
            interventions_module[name] = desc
        st.table(interventions_module)
    except KeyError:
        st.info("No interventions information available.")

    # Button to go to ClinicalTrials.gov and see the trial. It takes the user to the official page of the trial.
    st.markdown(
        f"See more in [ClinicalTrials.gov](https://clinicaltrials.gov/study/{trial['protocolSection']['identificationModule']['nctId']})"
    )


if __name__ == "__main__":
    username = "demo"
    password = "demo"
    hostname = os.getenv("IRIS_HOSTNAME", "localhost")
    port = "1972"
    namespace = "USER"
    CONNECTION_STRING = f"iris://{username}:{password}@{hostname}:{port}/{namespace}"

    try:
        engine = create_engine(CONNECTION_STRING)
        diseases = get_most_similar_diseases_from_uri("C1843013")
        for disease in diseases:
            print(disease)
    except Exception as e:
        print(e)

    try:
        print(get_uri_from_name(engine, "Alzheimer disease 3"))
    except Exception as e:
        print(e)

    clinical_record_info = get_clinical_records_by_ids(["NCT00841061"])
    print(clinical_record_info)

    textual_description = (
        "A disease that causes memory loss and other cognitive impairments."
    )
    encoder = SentenceTransformer("allenai-specter")
    diseases = get_diseases_related_to_a_textual_description(
        textual_description, encoder
    )
    for disease in diseases:
        print(disease)

    try:
        similarities = get_similarities_among_diseases_uris(
            [
                "http://identifiers.org/medgen/C4553765",
                "http://identifiers.org/medgen/C4553176",
                "http://identifiers.org/medgen/C4024935",
            ]
        )
        for similarity in similarities:
            print(
                f'{similarity[0].split("/")[-1]} and {similarity[1].split("/")[-1]} have a similarity of {similarity[2]}'
            )
    except Exception as e:
        print(e)

# %%