Run all queries for a drug and return structured results.

Run all queries for a gene and return structured results.

Renders conversation history from a list of messages, handling multiple tool calls.

Renders conversation history from a list of messages, handling multiple tool calls.

Retrieve k most similar examples.

Run a subprocess command with proper error handling. Args: cmd: Command and arguments to run. description: Human-readable descrip

Run a function with a timeout using threading instead of multiprocessing. This allows variables to persist in the global namespace between functio

Save a SimpleITK image to file. Args: image: SimpleITK Image object output_path: Path to save the image (mus

Search across preprint sources and merge results.

Search for GWAS studies associated with a disease or trait. Parameters ---------- disease_term : str Disease or trait name (

Compute standard summary statistics for a list of values.

Validate FDA API response structure. Args: response: The FDA API response dictionary required_fields: List of required top-l

Wait for server to be ready with retries.

Write CSV tables from results.

Build Essie fragments for biomarker expression requirements.

Build Essie fragment for brain metastases filter.

Create agent based on ollama

Ensure the URL section exists in the variant.

Find description from rdfs:comment, skos:definition, or IAO definition.

Find rdfs:label or skos:prefLabel text.

Format and add content to markdown. This function applies formatting to AI message content before adding it to the markdown. It proce

Generate a unique cache key for an API request. Args: endpoint: The API endpoint URL params: Query parameters Returns:

Radial Basis Function (RBF) / Squared Exponential Kernel. Computes covariance between points.

Build a complete drug result dict for report testing.

Make an API request with error handling and rate limiting.

Make API request with retry logic and error handling.

Helper to build minimal results list for report generation.

Parse search results from JSON strings.

Prepare resources for retrieval and return selected resource names. Args: prompt: The user's query Returns:

Query MyDisease.info for a disease name.

Query MyChem.info for a drug name.

Query MyGene.info for a gene symbol.

Core function to query NCBI databases using Claude for query interpretation and NCBI eutils. Parameters ---------- database (str): NCBI d

Core GET request with caching, rate limiting, and 429 retry.

Sort a list of association dicts by p-value (ascending, None last).

Retrieves contact and location details for a single clinical trial identified by its NCT ID. Parameters: - call_benefit: Define and

Retrieves outcome measures, results (if available), and adverse event data for a single clinical trial. Parameters: - call_benefit:

Retrieves publications and other references associated with a single clinical trial identified by its NCT ID. Parameters: - call_ben

Add a thought to the session.

Align short sequences (primers) to a longer sequence, allowing for one mismatch. Checks both forward and reverse complement strands. Args:

Assess replication of findings between discovery and replication studies. Args: tu: ToolUniverse instance trait: Disease or

Auto-detect format from question text and apply appropriate formatting. Args: value: Value to format question_text: Question text

Calculate CFD score for an off-target site CFD (Cutting Frequency Determination) predicts cleavage probability relative to the on-target site

Calculate Lipinski Rule of 5 properties.

Check if file exists and is readable.

Classify confidence score into categories.

Clear all sessions.

Compute per-population allele frequencies for the alt allele. Args: geno_matrix: (n_samples, n_variants), values 0/1/2/-1 pop_ind

Ratio of mean observed heterozygosity to theoretical max (0-1).

Run PCA on the genotype matrix. Handles missing data by mean imputation per variant. Returns: coords: (n_samples, n_components) prin

Measure how well sample proportions match global proportions (0-1). Returns: dict with keys: representation_index (float or None),

Count and aggregate results by a specific field. Args: category: API category endpoint: Specific endpoint

Count measurement documents in ASM.

Create SMARTS query from string or SMILES.

Detect outliers using Median Absolute Deviation (MAD). Parameters ---------- adata : AnnData Annotated data matrix with QC m

Detect outliers using Median Absolute Deviation (MAD). Parameters ---------- adata : AnnData Annotated data matrix with QC m

Discover paired alignment and tree files for a group. Args: data_dir: root directory to search group_name: optional subdirectory

Drafts a regulatory submission section using an agentic feedback loop. Args: section_name: e.g. "Pediatric Asses

Generate embeddings for semantic similarity.

Evaluate a user's answer for a given task and instance Args: task_name: Name of the task (e.g., 'gwas_causal_gene_openta

Extract drug names from an OpenFDA result.

Extract a row from a guideline annotation.

Fetch the text of a list of PubMed IDs.

Fetch detailed SRA run information using efetch CSV format. This provides richer metadata than esummary, including sample names. Args:

Filter out intronic and intergenic variants. Returns: (non_intronic_intergenic, intronic_intergenic)

Remove variants where all samples are homozygous reference (0/0 or ./.).

Filter out specified paths from variant data. Args: variants: List of variant dictionaries from MyVariant.info API Returns:

Find the best matching module from a list of available modules. This function attempts to match a target module name against a list of availa

Find enzymes that catalyze a specific transformation.

Flatten ASM JSON to list of row dictionaries. Each measurement becomes one row with metadata repeated.

Flatten nested list to 1D array of floats.

Format enhanced annotations for display.

Format intervention details as markdown. Args: intervention: Intervention data dictionary Returns: Formatted markdown s