### Quick Start Examples

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/human_protein_atlas_database/SKILL.md

These examples demonstrate basic usage for resolving Ensembl IDs and retrieving subcellular locations. All commands write JSON output to disk, defaulting to `/tmp/hpa_output.json` if `--output` is omitted.

```bash
# Map the ERBB2 gene symbol to its Ensembl ID
uv run scripts/hpa_cli.py resolve-ensembl-id ERBB2 --output /tmp/erbb2_id.json

# Get subcellular location by Ensembl ID
uv run scripts/hpa_cli.py get-subcellular-location ENSG00000141736 --output /tmp/erbb2_location.json
```

--------------------------------

### Quick Start: Search BRCA1 Variants

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/clinvar_database/SKILL.md

Example of how to use the wrapper script to search for BRCA1 variants and save the results to a JSON file.

```bash
uv run scripts/clinvar_api.py search --query "BRCA1[gene]" --output results.json
```

--------------------------------

### Example Usage

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pubmed_database/SKILL.md

Examples demonstrating how to use the PubMed API CLI for searching and fetching article abstracts, and processing the JSON output with `jq`.

```APIDOC
## Example Usage

```bash
uv run scripts/pubmed_api.py ./search_results.json search_pubmed "BRCA1" --max_results 5
cat ./search_results.json | jq '.[]' -r
uv run scripts/pubmed_api.py ./abstracts.json fetch_article_abstracts "35113657"
cat ./abstracts.json | jq '.[0].title' -r
```
```

--------------------------------

### Install Science Skills Bundle

Source: https://github.com/google-deepmind/science-skills/blob/main/README.md

Use this command to install the Science Skills bundle via npx. This command adds the collection of agent skills for scientific research tasks.

```bash
npx skills add google-deepmind/science-skills/
```

--------------------------------

### Check uv Installation

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/uv/SKILL.md

Verify if the uv package manager is installed and accessible on your system.

```bash
uv --version
```

--------------------------------

### Check uv Installation (Default Location)

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/uv/SKILL.md

Check if uv is installed in the default location but not yet on the system's PATH.

```bash
"$HOME/.local/bin/uv" --version
```

--------------------------------

### Workflow Example: Downloading Compound Structures

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/chembl_database/SKILL.md

This snippet shows how to download compound structures in SDF format.

```APIDOC
## Download Compound Structures

```bash
uv run scripts/chembl_api.py molecule --output /tmp/molecule.sdf --id CHEMBL1185377 --dl_format sdf
```

### Description
This command downloads the structure file for the molecule with ChEMBL ID `CHEMBL1185377` in SDF format. The downloaded file will be saved as `/tmp/molecule.sdf`. This demonstrates how to retrieve structural data for compounds.
```

--------------------------------

### Loading Structures

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pymol/references/PYMOL_REFERENCE.md

Examples of how to load molecular structures from various file formats into PyMOL.

```APIDOC
### Loading structures

```python
cmd.load("data/structure.cif", "myprotein")
cmd.load("data/structure.pdb", "myprotein")
```
```

--------------------------------

### FASTA Protein Sequence Input Example

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/protein_sequence_msa/SKILL.md

Input must be a plain text file with two or more protein sequences in FASTA format. Each sequence header must start with a '>' symbol.

```plaintext
>Sequence_1_Name
MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQ
QRLIFAGKQLEDGRTLSDYNIQKESTLHLVLRLRGG
>Sequence_2_Name
MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQ
QRLIFAGKQLEDGRTLSDYNIQKESTLHLVLRLRGG
```

--------------------------------

### Install uv Package Manager

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/uv/SKILL.md

Installs the uv package manager using a curl script. This command should be run if uv is not detected on the system.

```bash
curl -LsSf https://astral.sh/uv/install.sh | sh
```

--------------------------------

### Quick Start: Map Gene Symbol and Get Median Expression

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/gtex_database/SKILL.md

Example commands to quickly map a gene symbol to its GENCODE ID and then retrieve its median expression levels. Output files are saved in the /tmp/ directory.

```bash
# Map the TNF gene symbol to its GENCODE ID
uv run scripts/gtex_cli.py resolve-gencode-id TNF --output /tmp/tnf_id.json

# Get median expression of a gene by GENCODE ID
uv run scripts/gtex_cli.py get-median-expression ENSG00000232810.2 --output /tmp/tnf_expr.json
```

--------------------------------

### Search PubMed Example

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pubmed_database/SKILL.md

Example of how to search PubMed for a query and save the results to a JSON file. The results can then be processed with jq.

```bash
uv run scripts/pubmed_api.py ./search_results.json search_pubmed "BRCA1" --max_results 5
```

```bash
cat ./search_results.json | jq '.[]' -r
```

--------------------------------

### Workflow Example: Verifying API Status

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/chembl_database/SKILL.md

This snippet demonstrates how to verify if the ChEMBL API is available by running the `status` subcommand.

```APIDOC
## Verify API Status

```bash
uv run scripts/chembl_api.py status --output /tmp/status.json
```

### Description
This command checks the availability of the ChEMBL API. It runs the `status` subcommand and saves the output to `/tmp/status.json`. This is often the first step in a workflow to ensure the API is operational before proceeding with other queries.
```

--------------------------------

### Search for Molecules

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/chembl_database/SKILL.md

Example of searching for molecules using the 'molecule' subcommand. Requires an output file and a search query.

```bash
molecule --search "aspirin" --output /tmp/aspirin.json
```

--------------------------------

### Multi-step Query: Search and Get Relations

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/embl_ebi_ols/SKILL.md

Perform a two-step query to find a term's OBO ID and then retrieve its relations. First, search for the term in a specific ontology, saving results to a temporary file. Then, use the retrieved OBO ID to get relations.

```bash
search_ols.py --query "myocardial infarction" --ontology doid --exact --rows 1 --output /tmp/step1.json
```

```bash
get_term.py --obo_id DOID:5844 --relations parents --output /tmp/step2.json
```

--------------------------------

### Paginated List Endpoint Example

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/chembl_database/SKILL.md

Demonstrates how to use --limit and --offset for paginating results from list endpoints.

```APIDOC
## Paginated List Endpoint Example

### Description
All list endpoints support `--limit` and `--offset` for pagination. The response includes `page_meta` with pagination details.

### Method
`uv run scripts/chembl_api.py <list_endpoint>`

### Parameters
#### Query Parameters
- **--limit** (integer) - Optional - The number of results per page.
- **--offset** (integer) - Optional - The starting offset for the results.
- **--output** (string) - Required - The path to save the output JSON file.

### Request Examples
```bash
# First page: 2 results starting at offset 0
uv run scripts/chembl_api.py molecule --limit 2 --offset 0 --output /tmp/page1.json

# Second page: next 2 results starting at offset 2
uv run scripts/chembl_api.py molecule --limit 2 --offset 2 --output /tmp/page2.json
```
```

--------------------------------

### Workflow Example: Searching for Targets

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/chembl_database/SKILL.md

This snippet shows how to search for drug targets using the `target` subcommand and common options.

```APIDOC
## Search for Targets

```bash
uv run scripts/chembl_api.py target --output /tmp/targets.json --search "BRCA1"
```

### Description
This command searches for targets related to "BRCA1" using the `target` subcommand. The results will be saved to `/tmp/targets.json`. This illustrates a typical search operation for a specific entity.
```

--------------------------------

### Fetch Article Abstracts Example

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pubmed_database/SKILL.md

Example of how to fetch article abstracts using their PubMed IDs and save them to a JSON file. The titles can then be extracted using jq.

```bash
uv run scripts/pubmed_api.py ./abstracts.json fetch_article_abstracts "35113657"
```

```bash
cat ./abstracts.json | jq '.[0].title' -r
```

--------------------------------

### Import http_client from science_skills_common

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/science_skills_common/SKILL.md

Skills should import the http_client utility from the science_skills_common package. This package is automatically installed as a dependency.

```python
from science_skills.science_skills_common import http_client
```

--------------------------------

### Get References for an Article

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/literature_search_europepmc/SKILL.md

Fetches the reference list (bibliography) for a given article. Specify the source database and article ID.

```bash
uv run scripts/europepmc_api.py get_references MED 34265844 \
  --page_size 100 --output references.json
```

--------------------------------

### Workflow Example: Fetching Detailed Records

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/chembl_database/SKILL.md

This snippet demonstrates how to fetch detailed information for a specific target using its ID.

```APIDOC
## Fetch Detailed Target Record

```bash
uv run scripts/chembl_api.py target --output /tmp/target_details.json --id CHEMBL3893
```

### Description
This command fetches the detailed record for a specific target identified by the ChEMBL ID `CHEMBL3893`. The detailed information will be saved to `/tmp/target_details.json`. This shows how to retrieve specific records once an ID is known.
```

--------------------------------

### Workflow Example: Querying Bioactivity Data

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/chembl_database/SKILL.md

This snippet illustrates how to query bioactivity data for a specific target, with options for normalization.

```APIDOC
## Query Bioactivity Data

```bash
uv run scripts/chembl_api.py activity --output /tmp/activities.json --id CHEMBL3893 --normalize
```

### Description
This command queries the bioactivity data associated with the target `CHEMBL3893`. The `--normalize` option ensures that the activity values are converted to nM for consistent comparison across different studies. The results are saved to `/tmp/activities.json`.
```

--------------------------------

### Mandatory Initialization Boilerplate

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pymol/references/PYMOL_REFERENCE.md

Every PyMOL script must start with this exact sequence. The order matters for correct initialization.

```APIDOC
## Mandatory Initialization Boilerplate

Every PyMOL script **must** start with this exact sequence. The order matters —
reversing the import and `finish_launching()` will crash.

```python
import pymol
pymol.pymol_argv = ["pymol", "-cq"]
pymol.finish_launching()
from pymol import cmd
```

-   `-c` = command-line mode (no GUI)
-   `-q` = quiet (suppress startup messages)
```

--------------------------------

### Mandatory PyMOL Initialization

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pymol/references/PYMOL_REFERENCE.md

This boilerplate code must start every PyMOL script. The order of import and finish_launching is critical for correct execution.

```python
import pymol
pymol.pymol_argv = ["pymol", "-cq"]
pymol.finish_launching()
from pymol import cmd
```

--------------------------------

### Set up openFDA API Key

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/openfda_database/SKILL.md

Use this command to securely add your openFDA API key to the .env file. This is recommended to increase daily request limits.

```bash
printf "Enter openFDA API key (typing hidden): " && read -s key && echo && echo "FDA_API_KEY=$key" >> "ENV_FILE" && echo "Saved."
```

--------------------------------

### Get Ontology Metadata

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/embl_ebi_ols/SKILL.md

Retrieve metadata for a specific ontology using its ID. For example, to get information about EFO.

```bash
get_ontology.py --id <id>
```

--------------------------------

### Get Autocomplete Suggestions for Terms

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/embl_ebi_ols/SKILL.md

This script provides autocomplete suggestions for partial ontology term names. Results are saved to a specified JSON file.

```bash
uv run scripts/suggest_ols.py --query "diabet" --rows 5 \
  --output /tmp/ols_suggest.json 2>/dev/null
```

--------------------------------

### Get Canonical Transcription Start Site (TSS)

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/ensembl_database/SKILL.md

Obtains the single coordinate for the Transcription Start Site (TSS) of a gene's canonical transcript. Accepts gene symbols or Ensembl IDs and handles strand orientation.

```bash
uv run scripts/ensembl_api.py canonical-tss TP53 --output tp53_tss.json
```

```bash
uv run scripts/ensembl_api.py canonical-tss ENSG00000141510 --output tss.json
```

--------------------------------

### Setup AlphaGenome DNA Client

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/alphagenome_single_variant_analysis/SKILL.md

Initializes the AlphaGenome DNA client using an API key from environment variables. Ensure the ALPHAGENOME_API_KEY is set before running.

```python
from alphagenome.models import dna_client
from alphagenome.models import variant_scorers
from alphagenome.data import genome
import os
import pandas as pd

# Setup API Key and Client
dna_model = dna_client.create(api_key=os.environ.get('ALPHAGENOME_API_KEY'),
                              address='dns:///gdmscience.googleapis.com:443')
```

--------------------------------

### Minimal PyMOL Script for Rendering Structures

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pymol/SKILL.md

This script demonstrates the basic setup and commands for loading a structure, applying basic coloring, and saving a PNG image and session file. Ensure structure files are downloaded and accessible. Always call cmd.quit() at the end.

```python
# /// script
# requires-python = ">=3.10, <3.13"
# dependencies = [
#     "pymol-open-source-whl",
# ]
# ///

import os
import sys

# Set environment variable for headless rendering
os.environ["PYOPENGL_PLATFORM"] = "osmesa"

import pymol # pytype: disable=import-error
pymol.pymol_argv = ["pymol", "-cq"]
pymol.finish_launching()

from pymol import cmd # pytype: disable=import-error

cmd.load("AF-P00520-F1-model_v4.cif", "structure")
cmd.show("cartoon")
cmd.color("green", "ss h")
cmd.color("yellow", "ss s")
cmd.color("gray", "ss l+''")
cmd.orient()
cmd.set("ray_opaque_background", 1)
cmd.png("output/render.png", width=1200, height=900, dpi=150)
cmd.save("output/session.pse")
cmd.quit()

```

--------------------------------

### Get ENTEx Data for Genomic Region

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/encode_ccres_database/SKILL.md

Use the 'entex' command to retrieve ENTEx data for a specified genomic region. Requires region coordinates in 'chr:start:end' format.

```bash
uv run scripts/screen_api.py entex \
    --region chr1:1000068:1000409 \
    --output /tmp/entex.json
```

--------------------------------

### Set up NCBI API Key

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/dbsnp_database/SKILL.md

Use this command to securely set your NCBI API key in the `.env` file. It prompts for the key and appends it to the specified environment file without exposing it in the chat context.

```bash
printf "Enter NCBI API key (typing hidden): " && read -s key && echo && echo "NCBI_API_KEY=$key" >> "ENV_FILE" && echo "Saved."
```

--------------------------------

### Get eQTLs in Chromosomal Region

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/gtex_database/SKILL.md

Fetch significant single-tissue eQTLs within a specified chromosomal range. Supports a maximum window of 8Mb and requires chromosome, start, end, and tissue ID.

```bash
uv run scripts/gtex_cli.py get-eqtls-in-region chr17 7000000 7100000 "Esophagus - Muscularis" \
  --output /tmp/region_eqtls.json
```

--------------------------------

### Exploratory Query with CLI

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/interpro_database/SKILL.md

Use the CLI for exploratory queries with a strict limit and output to a file. This prevents overwhelming the context window and fetching excessive data.

```bash
uv run ./scripts/interpro_client.py fetch protein --source_db reviewed --limit 2 --query_params tax_id=9606 --output exploratory_results.jsonl
```

--------------------------------

### Initialize AlphaGenome DNA Client

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/alphagenome_single_variant_analysis/docs/alphagenome-api.md

Create an instance of the DNA client. The API key is loaded from the .env file.

```python
from alphagenome.models import dna_client

dna_model = dna_client.create(
    api_key=os.environ.get('ALPHAGENOME_API_KEY'),
    address='dns:///gdmscience.googleapis.com:443',
)
```

--------------------------------

### Search with Pagination Workflow

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/literature_search_europepmc/SKILL.md

Illustrates how to perform searches with pagination using the `cursor` parameter. Requires `jq` for extracting cursor information.

```bash
# First page
uv run scripts/europepmc_api.py search "CRISPR" --max_results 100 --output page1.json
# Extract cursor for next page
CURSOR=$(jq -r '.nextCursorMark // empty' page1.json)
# Next page
uv run scripts/europepmc_api.py search "CRISPR" --max_results 100 --cursor "$CURSOR" --output page2.json
```

--------------------------------

### Get API Key

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/string_database/references/valuesranks.md

Generate a free, anonymous API key required to submit full datasets. Read the JSON output to get the "api_key" value.

```bash
uv run scripts/string_cli.py valuesranks-key --output /tmp/api_key.json
```

--------------------------------

### Fetch PDB Structures for an Entry via CLI

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/interpro_database/SKILL.md

Retrieve PDB structures linked to an InterPro entry. This example limits the fetch to the first 5 structures and saves them to a JSONL file.

```bash
# URL equivalent: /structure/pdb/entry/interpro/IPR011615
# Only fetch the first 5 structures
uv run ./scripts/interpro_client.py fetch structure
    --source_db pdb
    --linked_endpoint entry
    --linked_source_db interpro
    --linked_accession IPR011615
    --output ipr011615_structures.jsonl
```

--------------------------------

### Fetch Entries (/{endpoint})

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/interpro_database/SKILL.md

Construct the endpoint for fetching entries directly using the `/{endpoint}` path. Ensure to use a limit for exploratory queries.

```bash
uv run ./scripts/interpro_client.py fetch entry --limit 10 --output entries.jsonl
```

--------------------------------

### Get Subcellular Location

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/human_protein_atlas_database/SKILL.md

Retrieves the specific organelles or cellular structures where the protein has been localized.

```APIDOC
## get-subcellular-location — Get Subcellular Location

Retrieves the specific organelles or cellular structures where the protein has been localized.

### Command
```bash
uv run scripts/hpa_cli.py get-subcellular-location ENSG00000141736 \
  --output /tmp/subcellular.json
```

### Arguments
- `ensembl_id` (positional): The Ensembl Gene ID.
- `--output`: Output file path.
```

--------------------------------

### Execute BLAST Search (Default Database)

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/protein_sequence_similarity_search/SKILL.md

Run the uniprot_blast.py script with the default 'uniprotkb' database. Ensure the USER_EMAIL environment variable is set.

```bash
uv run scripts/uniprot_blast.py <SEQUENCE_OR_FILE> -o <generated-filename.md> -j <generated-filename.json>
```

--------------------------------

### Get OLS Index Statistics

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/embl_ebi_ols/SKILL.md

Retrieve statistics about the OLS index. This command does not require any arguments.

```bash
get_stats.py
```

--------------------------------

### Coloring

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pymol/references/PYMOL_REFERENCE.md

Examples of applying different color schemes to molecular structures based on selections or properties.

```APIDOC
### Coloring

```python
cmd.color("green", "ss h")
cmd.color("cyan", "chain A")
cmd.spectrum("b", "red_white_blue", "polymer.protein")
cmd.spectrum("count", "rainbow", "polymer.protein")
```
```

--------------------------------

### Standard AlphaGenome Imports

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/alphagenome_single_variant_analysis/docs/alphagenome-api.md

Import necessary modules for AlphaGenome workflows. Ensure all required packages are installed.

```python
from alphagenome.data import gene_annotation
from alphagenome.data import genome
from alphagenome.data import track_data
from alphagenome.data import transcript as transcript_utils
from alphagenome.interpretation import ism
from alphagenome.models import dna_client
from alphagenome.models import variant_scorers
from alphagenome.visualization import plot_components
import matplotlib.pyplot as plt
import pandas as pd
```

--------------------------------

### Run openFDA Utility Script

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/openfda_database/SKILL.md

This is the main command to interact with the openFDA API using the provided utility script. It supports search, count, and download operations.

```bash
uv run scripts/openfda_query.py {search,count,download} --output <file> [options]
```

--------------------------------

### Get Reactome Database Name

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/reactome_database/SKILL.md

Retrieves the name of the Reactome database. Results are saved to a JSON file.

```bash
uv run scripts/reactome_analysis.py db-name --output /tmp/name.json
```

--------------------------------

### Get Protein Information

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/ensembl_database/SKILL.md

Retrieves the Ensembl Protein ID (ENSP) and sequence length for a given transcript.

```bash
uv run scripts/ensembl_api.py protein-info ENST00000269305 --output protein_info.json
```

--------------------------------

### Fetch Compound Database Summary

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pubmed_database/references/fetch-and-resolve.md

Use this command to fetch summary metadata for compound records from the pccompound database. Provide a single UID.

```bash
uv run scripts/pubmed_api.py ./compound_summary.json fetch_database_summary pccompound "2244"
```

--------------------------------

### Get Ontology Root Terms

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/embl_ebi_ols/SKILL.md

Retrieve the root terms of a specified ontology. Requires the ontology ID.

```bash
get_term.py --ontology <id> --roots
```

--------------------------------

### Client Initialization

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/alphagenome_single_variant_analysis/docs/alphagenome-api.md

Initializes the AlphaGenome DNA client. The API key is automatically loaded from the .env file.

```APIDOC
## Client Initialization

### Description
Initializes the AlphaGenome DNA client. The API key is automatically loaded by `dotenv` from the `.env` file in the agent configuration directory.

### Code
```python
from alphagenome.models import dna_client

dna_model = dna_client.create(
    api_key=os.environ.get('ALPHAGENOME_API_KEY'),
    address='dns:///gdmscience.googleapis.com:443',
)
```
```

--------------------------------

### List Searchable Field Areas

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/clinical_trials_database/references/clinical_trials_api.md

Get a list of all areas or categories that can be searched within the study data.

```APIDOC
## GET /studies/search-areas

### Description
Lists all searchable field areas.

### Method
GET

### Endpoint
/studies/search-areas
```

--------------------------------

### Fetch Entries by Source DB (/{endpoint}/{sourceDB})

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/interpro_database/SKILL.md

Fetch entries filtered by a specific source database using the `/{endpoint}/{sourceDB}` path. Use a limit for exploratory queries.

```bash
uv run ./scripts/interpro_client.py fetch entry --source_db pfam --limit 10 --output pfam_entries.jsonl
```

--------------------------------

### Create GO Slim Summary from Gene Annotations

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/quickgo_database/SKILL.md

Use this command to generate a high-level functional summary by mapping gene annotations to a predefined GO Slim. First, obtain GO IDs for your genes, then use the `go slim` command with the relevant GO IDs.

```bash
# Step 1: Find GO IDs for candidate genes (e.g., via their UniProt IDs, fetching their annotations)
# ... (output yields e.g., GO:0006915,GO:0008219)

# Step 2: Create a slim summary from those specific GO IDs
uv run scripts/quickgo_tool.py go slim --slimsToIds "GO:0005575,GO:0008150,GO:0003674" --slimsFromIds "GO:0006915,GO:0008219" --output my_slim.json
```

--------------------------------

### Run STRING CLI Tool

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/string_database/SKILL.md

This is the general command structure for running the STRING CLI tool. Always use 'uv run' and specify an output file.

```bash
uv run scripts/string_cli.py <command> [options] --output /tmp/out.tsv
```

--------------------------------

### Get Reactome Database Version

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/reactome_database/SKILL.md

Retrieves the current version of the Reactome database. Results are saved to a JSON file.

```bash
uv run scripts/reactome_analysis.py db-version --output /tmp/version.json
```

--------------------------------

### Get Dataset Details

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/unibind_database/SKILL.md

Retrieves detailed information for a specific dataset in UniBind using its identifier. Output is JSON.

```bash
uv run <SKILL DIR>/scripts/unibind_api.py get_dataset "EXP047889.HMLE-Twist-ER_breast_cancer.SMAD3"
```

--------------------------------

### Download All Results (Auto-Paginate)

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/openfda_database/references/recipes.md

Downloads all results for a given query, automatically handling pagination. Useful for retrieving large datasets.

```bash
uv run scripts/openfda_query.py download \
  --category drug --endpoint event \
  --search "patient.drug.medicinalproduct:ozempic+AND+serious:1" \
  --limit 100 --all_results \
  --output /tmp/ozempic_serious.json
```

--------------------------------

### Get Assembly/Region Info

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/ensembl_database/references/ensembl_rest_api_reference.md

Retrieves details for a specific chromosome or genomic region, including its length and cytogenetic bands.

```APIDOC
## GET /info/assembly/{species}/{region_name}

### Description
Details for a specific chromosome/region (length, bands).

### Method
GET

### Endpoint
/info/assembly/{species}/{region_name}

### Parameters
#### Path Parameters
- **species** (string) - Required - The species name (e.g., 'human').
- **region_name** (string) - Required - The name of the region (e.g., '1:1000000..2000000').
```

--------------------------------

### Visualize Protein-Ligand Interactions

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pymol/references/RECIPES.md

Loads a complex, isolates a ligand, selects the binding site, and styles the protein and ligand for visualization. It also shows pocket waters and polar contacts, then zooms and saves a PNG and session file. Raises an error if no ligand is found.

```python
cmd.load("data/complex.pdb", "complex")

# Isolate a single ligand to prevent zoomed-out views on symmetrical complexes
if cmd.count_atoms("organic") > 0:
    first_atom = cmd.get_model("organic").atom[0]
    cmd.select("target_ligand", f"organic and chain '{first_atom.chain}'")
else:
    raise RuntimeError("No ligand found.")

cmd.select("binding_site", "byres (polymer.protein within 4.0 of target_ligand)")

# Styled rendering with heteroatom coloring
cmd.show("cartoon", "polymer.protein")
cmd.set("cartoon_transparency", 0.4, "polymer.protein")
cmd.color("gray80", "polymer.protein")

cmd.show("sticks", "binding_site")
util.cbac("binding_site")
util.cnc("binding_site")
cmd.show("sticks", "target_ligand")
util.cbag("target_ligand")
util.cnc("target_ligand")
cmd.hide("sticks", "hydro")

# Show pocket waters
cmd.select("pocket_waters", "solvent within 4.0 of target_ligand")
cmd.show("spheres", "pocket_waters")
cmd.color("red", "pocket_waters")
cmd.set("sphere_scale", 0.15, "pocket_waters")

# Polar contacts (hydrogen bonds)
cmd.distance("polar_contacts", "target_ligand", "binding_site", cutoff=3.5, mode=2)
cmd.set("dash_color", "yellow")
cmd.set("dash_gap", 0.4)
cmd.set("dash_radius", 0.08)
cmd.hide("labels", "polar_contacts")

cmd.orient("target_ligand | binding_site")
cmd.zoom("target_ligand | binding_site", buffer=3.0)
cmd.png("output/ligand.png", width=1200, height=900, dpi=150)
cmd.save("output/session.pse")
```

--------------------------------

### Execute BLAST Search (Custom Database)

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/protein_sequence_similarity_search/SKILL.md

Run the uniprot_blast.py script specifying custom databases. Ensure the USER_EMAIL environment variable is set.

```bash
uv run scripts/uniprot_blast.py <SEQUENCE_OR_FILE> -o <generated-filename.md> -j <generated-filename.json> --databases <db1,db2>
```

--------------------------------

### Get Gene Summary

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/ensembl_database/SKILL.md

Fetches high-level gene metadata including symbol, biotype, description, and chromosomal location.

```bash
uv run scripts/ensembl_api.py gene-summary ENSG00000141510 --output gene_summary.json
```

--------------------------------

### Get Hierarchical Children

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/embl_ebi_ols/SKILL.md

Retrieve hierarchical children (including 'is-a' and 'part-of' relations) for a given ontology term.

```bash
get_term.py --obo_id <ID> --relations hierarchicalChildren
```

--------------------------------

### Pagination Query Syntax

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/openfda_database/references/api_endpoints.md

Explains the `limit` and `skip` parameters for controlling the number of results returned and for paginating through large result sets.

```text
limit=N
skip=N
```

--------------------------------

### Execute SPARQL Queries using uniprot_tools.py

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/uniprot_database/SKILL.md

Run complex graph queries directly against UniProt using SPARQL via the `sparql` command in `uniprot_tools.py`. This example retrieves the first 5 reviewed proteins.

```bash
uv run scripts/uniprot_tools.py sparql 'PREFIX up: <http://purl.uniprot.org/core/> SELECT ?protein WHERE { ?protein a up:Protein ; up:reviewed true . } LIMIT 5'
```

--------------------------------

### Get Hierarchical Parents

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/embl_ebi_ols/SKILL.md

Retrieve hierarchical parents (including 'is-a' and 'part-of' relations) for a given ontology term.

```bash
get_term.py --obo_id <ID> --relations hierarchicalParents
```

--------------------------------

### Rendering Backend

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/pymol/references/PYMOL_REFERENCE.md

Information about the PyMOL rendering backend and recommended functions for output.

```APIDOC
## Rendering Backend

PyMOL runs with **OSMesa** (software rendering). There is no GPU or X display.

-   Use `cmd.png(path, width, height, dpi)` for output.
-   `cmd.ray()` works but is slow — use it only when you need ray-traced
    quality.
-   Never use `cmd.draw()` (requires hardware OpenGL).
-   Always set `cmd.set("ray_opaque_background", 1)` if you want a white
    background instead of transparent.
```

--------------------------------

### Get Protein Sequence

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/ensembl_database/SKILL.md

Fetches the amino acid sequence in FASTA format for a transcript (ENST) or protein (ENSP) ID.

```bash
uv run scripts/ensembl_api.py protein-sequence ENST00000269305 --output protein.fasta
```

```bash
uv run scripts/ensembl_api.py protein-sequence ENSP00000269305 --output protein_ensp.fasta
```

--------------------------------

### Fetch Entries by Source DB and Accession (/{endpoint}/{sourceDB}/{accession})

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/interpro_database/SKILL.md

Retrieve a specific entry by its source database and accession number using the `/{endpoint}/{sourceDB}/{accession}` path. A limit is recommended for exploratory queries.

```bash
uv run ./scripts/interpro_client.py fetch entry --source_db pfam --accession PF00001 --limit 10 --output pf00001_entry.jsonl
```

--------------------------------

### Get Transcript Structure

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/ensembl_database/SKILL.md

Fetches exon coordinates, CDS boundaries, and computed 5'/3' UTR regions for a given transcript.

```bash
uv run scripts/ensembl_api.py transcript-structure ENST00000269305 --output structure.json
```

--------------------------------

### Get Term Ancestors

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/embl_ebi_ols/SKILL.md

Retrieve all ancestors (categories) of a given ontology term. Use the --relations ancestors flag.

```bash
get_term.py --obo_id <ID> --relations ancestors
```

--------------------------------

### Paginate Molecule Results

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/chembl_database/SKILL.md

Demonstrates how to paginate through molecule results using the --limit and --offset parameters. This allows fetching data in chunks.

```bash
# First page: 2 results starting at offset 0
uv run scripts/chembl_api.py molecule --limit 2 --offset 0 --output /tmp/page1.json

# Second page: next 2 results starting at offset 2
uv run scripts/chembl_api.py molecule --limit 2 --offset 2 --output /tmp/page2.json
```

--------------------------------

### Get Term Parents

Source: https://github.com/google-deepmind/science-skills/blob/main/skills/embl_ebi_ols/SKILL.md

Retrieve the direct parents of a given ontology term. Use the --relations parents flag.

```bash
get_term.py --obo_id <ID> --relations parents
```