# REST API drugs by gene: TP53 to drug potency

> Script a gene-to-drug lookup with the BioQ REST API: one call for TP53 returns 18 drugs with stage and pChEMBL potency, data as of 2026-11-06.

Data as of 2026-11-06 · Published 2026-11-06 · Updated 2026-11-06 · https://bioq.bio/blog/en/api-recipe-every-drug-for-a-gene-over-rest

![BioQ search page with the query TP53 and a table of drugs ordered by clinical stage, with mechanism and pChEMBL columns.](https://bioq.bio/blog/img/api-recipe-every-drug-for-a-gene-over-rest/en/hero.webp)

A single REST API call for drugs by gene turns the symbol TP53 into 18 known drugs, each with its maximum clinical stage and, where one exists, an exact pChEMBL potency. A second call adds CRISPR dependency across 1178 DepMap cell lines. The two calls took 0.21 engine seconds in total, and the data are as of 2026-11-06.

## What we asked: REST API drugs by gene for TP53

The question is the one a target-assessment script usually starts with: for this gene symbol, which drugs are known, how far did they get clinically, and how potent are they against the target? A second question follows from it: does the gene matter for cell viability in a knockout screen, and in which lineages?

Done by hand, this means a target lookup in Open Targets, a potency search in ChEMBL, a trial search in ClinicalTrials.gov and a DepMap download. Each has its own identifiers and its own export format. The BioQ REST API puts the first two questions behind two routes, `drugs` and `depmap`. Both take the gene symbol as a parameter, and both can be called from Python, from curl, or by an AI assistant over MCP.

## What BioQ returned from the drugs and depmap routes

The first route is `drugs`. The parameters are `gene` and `limit`. We set the gene to TP53 and the limit to 8. The response says 18 drugs are known for the gene and that 8 are shown. The source release combines Open Targets, ChEMBL, the Drug Repurposing Hub and ClinicalTrials.gov.

```bash
curl -s -H "Authorization: Bearer $BIOQ_API_KEY" \
  "https://bioq.bio/api/v1/drugs?gene=TP53&limit=8"
```

Each row carries the drug name, the maximum clinical stage, the mechanism, the exact pChEMBL value with its assay type, and a count of trials found by drug name.

| Drug | Max clinical stage | Mechanism | Best exact pChEMBL | Type | Trials by name |
|---|---|---|---|---|---|
| Aspirin | APPROVAL | — | — | — | 533 |
| Navtemadlin | PHASE_3 | Tumour suppressor p53/oncoprotein Mdm2 inhibitor | 9.22 | IC50 | 9 |
| Idasanutlin | PHASE_3 | Tumour suppressor p53/oncoprotein Mdm2 inhibitor | 8.22 | IC50 | 14 |
| Contusugene Ladenovec | PHASE_3 | Cellular tumor antigen p53 exogenous gene | — | — | 0 |
| Eprenetapopt | PHASE_3 | Cellular tumor antigen p53 stabiliser | — | — | 1 |
| Teprasiran | PHASE_3 | p53 mRNA RNAi inhibitor | — | — | 1 |
| Alrizomadlin | PHASE_2 | Tumour suppressor p53/oncoprotein Mdm2 inhibitor | — | — | 1 |
| Cenersen | PHASE_2 | p53 mRNA antisense inhibitor | — | — | 3 |

The second route is `depmap`. The parameters are `gene` and `top`. We set the gene to TP53 and `top` to 6, which returns the six lineages listed. The response also carries gene-level scalars: 1178 cell lines, 4 of them dependent, which is 0.3 percent, and a mean gene effect of 0.379. The field `common_essential` is false.

```bash
curl -s -H "Authorization: Bearer $BIOQ_API_KEY" \
  "https://bioq.bio/api/v1/depmap?gene=TP53&top=6"
```

| Lineage | Cell lines | Dependent | % dependent | Mean gene effect |
|---|---|---|---|---|
| Head and Neck | 77 | 1 | 1.3 | 0.133 |
| Breast | 53 | 1 | 1.9 | 0.176 |
| Pancreas | 47 | 0 | 0 | 0.193 |
| Myeloid | 42 | 1 | 2.4 | 0.194 |
| Esophagus/Stomach | 69 | 0 | 0 | 0.223 |
| Lung | 126 | 0 | 0 | 0.226 |

The two queries drew on 2 distinct resources in the facts for this answer. The engine time was 0.21 seconds for both. That is the measured time on the BioQ side. It does not include your network latency or the time you spend writing the parsing code.

## What it means and how to read it

**Order of calls.** Call `drugs` first. It gives the list you will probably want to filter or join. Call `depmap` second if you need to know whether the gene is a viability dependency. The calls are independent, so a script can run them in parallel. The only thing they share is the gene symbol.

**Potency.** The pChEMBL value is the negative log10 of the molar potency, so a higher number means a stronger compound. In the table, Navtemadlin has 9.22 and Idasanutlin has 8.22, both from IC50 measurements. Both are listed with the mechanism "Tumour suppressor p53/oncoprotein Mdm2 inhibitor". This is the TP53 MDM2 axis. These compounds act on MDM2, the negative regulator of p53, and are listed under TP53 through that mechanism. They do not bind a mutant p53 directly. Keep that in mind when you filter by gene.

**Empty potency cells.** Several rows have no pChEMBL value and no assay type. That includes the gene therapy Contusugene Ladenovec, the RNAi drug Teprasiran and the antisense drug Cenersen. A potency number from a binding or enzyme assay does not apply to these modalities. In a script, treat the null as "no exact potency recorded", not as zero.

**Trial counts.** The last column counts trials found by drug name. Aspirin shows 533, with no mechanism and no potency recorded, and it appears with the stage APPROVAL. The count is a name match, not a count of TP53-related trials. A script that ranks drugs by this column will put aspirin on top for reasons unrelated to the gene.

**Dependency.** For TP53 the mean gene effect is 0.379. A more negative Chronos value means the knockout reduced viability more, so a positive mean means knockouts on average did not reduce viability in this panel. The 4 dependent lines are spread across Head and Neck, Breast and Myeloid, 1 each in the top six lineages shown. Pancreas, Esophagus/Stomach and Lung have 0 dependent lines. A dependency here means a gene effect at or below a named constant threshold. That threshold is a choice, not a property of the data.

**From an assistant.** The same two routes can be reached by an AI assistant over MCP. The assistant passes the same parameters and gets the same fields back, so the interpretation above applies unchanged. Keep the null handling and the name-match caveat in the instructions you give it.

## How to reproduce it in BioQ

1. Create an API key in your BioQ account and send it as a bearer token in the Authorization header of every request. For an MCP client, put the same key in the server configuration.
2. Call `drugs` with `gene=TP53` and `limit=8`, as in the first call above. In Python, pass the parameters as a dict to your HTTP client. In curl, add them to the query string.
3. Parse the rows by column name rather than position. Treat empty strings and nulls in the potency and type columns as missing values.
4. Call `depmap` with `gene=TP53` and `top=6`. Read the scalars for the gene-level summary and the rows for the lineage breakdown.
5. To check your script against the interface, run the same query in the BioQ search box and compare the tables:

```text
TP53
```

[Open this query in BioQ](https://bioq.bio/login?next=%2Fapp%23q%3DTP53)

6. Raise `limit` if you want all 18 drugs rather than the 8 shown, and raise `top` if you want more lineages.

## Limits: what the data do not show

- The `drugs` route lists drugs linked to the gene. It does not say that a drug works against TP53-mutant disease. Mechanism text and clinical stage come from curated sources and can differ in detail.
- pChEMBL values are exact measurements from single assay types, here IC50. They are not comparable across assay formats, and they say nothing about activity in patients.
- Trial counts by name can include trials with no link to the gene, as the aspirin row shows.
- DepMap covers cancer cell lines in culture. A positive mean gene effect for TP53 does not rule out dependence in a specific genetic context or in vivo. The breakdown by lineage uses small groups, so one dependent line can move a percentage a lot.
- The response is a snapshot as of 2026-11-06. Pin the date in your script if you need repeatable results, and expect the numbers to change as the underlying releases update.
- Nothing here is medical advice or a treatment recommendation.

## Data and method
The numbers on this page were read from the BioQ engine on 2026-11-06 (2 queries, 0.2 s of engine time in total). The tables are generated by code from the engine's answers; the text only restates them.

`/depmap`, `/drugs`

Run the same query yourself: [TP53](https://bioq.bio/login?next=%2Fapp%23q%3DTP53)

*For research use only. This page summarises public data and is not medical advice.*

BioQ is an AI search engine and agent with one goal: making biology research 100X faster. Ask in the web app, over the REST API, or from Claude, ChatGPT or Cursor through MCP.

## Queries used

- [`TP53`](https://bioq.bio/login?next=%2Fapp%23q%3DTP53)

## Frequently asked questions

### How do I get drugs by gene symbol from the BioQ REST API?

Call the `drugs` route with the parameter `gene` set to the symbol, for example TP53, and an optional `limit`. For TP53 the response reports 18 drugs and shows 8 with a limit of 8.

### What does pChEMBL mean in the drugs response?

It is the negative log10 of the molar potency, so a higher value means a stronger compound. For TP53, Navtemadlin has 9.22 and Idasanutlin 8.22, both from IC50 assays.

### Why do some TP53 drugs have no potency value?

Several entries, such as Cenersen, Teprasiran and Contusugene Ladenovec, are antisense, RNAi or gene therapy products. No exact pChEMBL is recorded for them, so the field is empty and should be treated as missing.

### Is TP53 a common essential gene in DepMap?

No. The `depmap` route returns common_essential as false. Of 1178 cell lines, 4 are dependent (0.3 percent), and the mean gene effect is 0.379.

### How long does the two-call recipe take?

The engine time was 0.21 seconds for 2 queries, drawing on 2 distinct resources. This excludes network latency and your own parsing time.

## Sources

- Dataset: [Open Targets Platform; ChEMBL; Drug Repurposing Hub](https://platform.opentargets.org/)
- Dataset: [DepMap CRISPR Chronos gene effect (24Q4)](https://depmap.org/portal/)

> Results come from public databases as of the date shown and may change. BioQ gives research information, not medical advice.
