cpic

Name: cpic
Author: QSong-github

132stars4forks

How to install cpic?

qsong-github/drugclaw/cpic

$npx -y skills add QSong-github/DrugClaw --skill cpic

Installs into the current project.

Summary

> Clinical Pharmacogenomics Implementation Consortium — gene-based prescribing guidelines > Category: Drug-centric | Type: DB | Subcategory: Drug Knowledgebase > API: https://api.cpicpgx.org/v1 (PostgREST, free, no key required)

Files · 1

View on GitHub

SKILL.md

1	# 67 · CPIC
2
3	> Clinical Pharmacogenomics Implementation Consortium — gene-based prescribing guidelines
4	> Category: Drug-centric \| Type: DB \| Subcategory: Drug Knowledgebase
5	> API: `https://api.cpicpgx.org/v1` (PostgREST, free, no key required)
6
7	\| Resource \| URL \|
8	\|----------\|-----\|
9	\| Homepage \| https://cpicpgx.org/ \|
10	\| API / Data \| https://cpicpgx.org/cpic-data/ \|
11	\| Paper \| https://pubmed.ncbi.nlm.nih.gov/33479744/ \|
12
13	---
14
15	## What it provides
16
17	- Drug metadata: drugid (RxNorm), DrugBank ID, ATC codes, flowchart links
18	- Guidelines: peer-reviewed pharmacogenomics prescribing guidelines (drug + gene → dosing advice)
19	- Gene-drug pairs: curated pairs with CPIC level, PharmGKB level, PGx testing status
20	- Dosing recommendations: phenotype-specific dosing adjustments per drug-gene combination
21
22	---
23
24	## API schema note
25
26	The `pair` and `recommendation` tables use `drugid` (e.g. `RxNorm:32968`), not drug name.
27	This script resolves drug names automatically via the `/v1/drug` table before querying.
28
29	Guideline lookup uses two strategies: (1) name substring match, (2) `guidelineid` from the drug table.
30	This is necessary because some guidelines use class names (e.g. simvastatin → `"SLCO1B1, ABCG2, CYP2C9, and Statins"`, codeine → `"CYP2D6, OPRM1, COMT, and Opioids"`).
31
32	---
33
34	## Quick start
35
36	```python
37	from 67_CPIC import query
38
39	# Single drug
40	results = query("clopidogrel")
41
42	# Multiple drugs
43	results = query(["warfarin", "codeine"])
44
45	# Query by gene symbol
46	results = query("CYP2D6", fields="pairs")
47
48	# Specific fields only
49	results = query("codeine", fields="guidelines")
50	results = query("codeine", fields="recommendations")
51	```
52
53	---
54
55	## `query()` interface
56
57	```
58	query(entities, fields="all") -> list[dict]
59	```
60
61	\| Parameter \| Type \| Description \|
62	\|------------\|--------------------\|-------------\|
63	\| `entities` \| `str \\| list[str]` \| Drug name(s) or gene symbol(s) \|
64	\| `fields` \| `str` \| `"all"` — everything; `"guidelines"` / `"pairs"` / `"recommendations"` \|
65
66	### Return structure (`fields="all"`)
67
68	```json
69	[
70	{
71	"query": "clopidogrel",
72	"drug_info": [
73	{"drugid": "RxNorm:32968", "name": "clopidogrel",
74	"drugbankid": "DB00758", "atcid": ["B01AC04"], "flowchart": "..."}
75	],
76	"guidelines": [
77	{"name": "CYP2C19 and Clopidogrel", "url": "...", "version": 66}
78	],
79	"gene_drug_pairs": [
80	{"genesymbol": "CYP2C19", "drugid": "RxNorm:32968",
81	"cpiclevel": "A", "clinpgxlevel": "1A",
82	"pgxtesting": "Actionable PGx", "citations": ["21716271", ...]}
83	],
84	"recommendations": [
85	{"drugid": "RxNorm:32968",
86	"phenotypes": {"CYP2C19": "Ultrarapid Metabolizer"},
87	"implications": {"CYP2C19": "Increased active metabolite ..."},
88	"recommendation": "Use at standard dose (75 mg/day)",
89	"classification": "Strong",
90	"population": "CVI ACS PCI"}
91	]
92	}
93	]
94	```
95
96	On error: `{"query": "xxx", "error": "..."}`.
97
98	---
99
100	## Lower-level functions
101
102	\| Function \| Input \| Output \| Description \|
103	\|----------\|-------\|--------\|-------------\|
104	\| `get_drug_info(drug_name)` \| drug name \| `list[dict]` \| Drug table lookup (fuzzy) \|
105	\| `get_guidelines(drug_name=None)` \| optional drug name \| `list[dict]` \| All or filtered guidelines \|
106	\| `get_gene_drug_pairs(drug_name=None, gene=None)` \| optional filters \| `list[dict]` \| Gene-drug pairs (name auto-resolved to drugid) \|
107	\| `get_recommendations(drug_name)` \| drug name \| `list[dict]` \| Dosing recommendations (name auto-resolved) \|
108
109	---
110
111	## Notes
112
113	- CPIC levels: A = guideline published, B = in progress, C/D = lower evidence.
114	- Gene symbols are auto-detected (uppercase, ≤12 chars) and routed to `genesymbol` filter.
115	- Drug names are fuzzy-matched via `ilike` on the `/v1/drug` table.
116	- No rate limit documented, but keep requests reasonable.

People who liked this also viewed

6 picks

Type

cpic

132stars4forks

View on GitHub

How to install cpic?

qsong-github/drugclaw/cpic

$npx -y skills add QSong-github/DrugClaw --skill cpic

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# 67 · CPIC
2
3	> Clinical Pharmacogenomics Implementation Consortium — gene-based prescribing guidelines
4	> Category: Drug-centric \| Type: DB \| Subcategory: Drug Knowledgebase
5	> API: `https://api.cpicpgx.org/v1` (PostgREST, free, no key required)
6
7	\| Resource \| URL \|
8	\|----------\|-----\|
9	\| Homepage \| https://cpicpgx.org/ \|
10	\| API / Data \| https://cpicpgx.org/cpic-data/ \|
11	\| Paper \| https://pubmed.ncbi.nlm.nih.gov/33479744/ \|
12
13	---
14
15	## What it provides
16
17	- Drug metadata: drugid (RxNorm), DrugBank ID, ATC codes, flowchart links
18	- Guidelines: peer-reviewed pharmacogenomics prescribing guidelines (drug + gene → dosing advice)
19	- Gene-drug pairs: curated pairs with CPIC level, PharmGKB level, PGx testing status
20	- Dosing recommendations: phenotype-specific dosing adjustments per drug-gene combination
21
22	---
23
24	## API schema note
25
26	The `pair` and `recommendation` tables use `drugid` (e.g. `RxNorm:32968`), not drug name.
27	This script resolves drug names automatically via the `/v1/drug` table before querying.
28
29	Guideline lookup uses two strategies: (1) name substring match, (2) `guidelineid` from the drug table.
30	This is necessary because some guidelines use class names (e.g. simvastatin → `"SLCO1B1, ABCG2, CYP2C9, and Statins"`, codeine → `"CYP2D6, OPRM1, COMT, and Opioids"`).
31
32	---
33
34	## Quick start
35
36	```python
37	from 67_CPIC import query
38
39	# Single drug
40	results = query("clopidogrel")
41
42	# Multiple drugs
43	results = query(["warfarin", "codeine"])
44
45	# Query by gene symbol
46	results = query("CYP2D6", fields="pairs")
47
48	# Specific fields only
49	results = query("codeine", fields="guidelines")
50	results = query("codeine", fields="recommendations")
51	```
52
53	---
54
55	## `query()` interface
56
57	```
58	query(entities, fields="all") -> list[dict]
59	```
60
61	\| Parameter \| Type \| Description \|
62	\|------------\|--------------------\|-------------\|
63	\| `entities` \| `str \\| list[str]` \| Drug name(s) or gene symbol(s) \|
64	\| `fields` \| `str` \| `"all"` — everything; `"guidelines"` / `"pairs"` / `"recommendations"` \|
65
66	### Return structure (`fields="all"`)
67
68	```json
69	[
70	{
71	"query": "clopidogrel",
72	"drug_info": [
73	{"drugid": "RxNorm:32968", "name": "clopidogrel",
74	"drugbankid": "DB00758", "atcid": ["B01AC04"], "flowchart": "..."}
75	],
76	"guidelines": [
77	{"name": "CYP2C19 and Clopidogrel", "url": "...", "version": 66}
78	],
79	"gene_drug_pairs": [
80	{"genesymbol": "CYP2C19", "drugid": "RxNorm:32968",
81	"cpiclevel": "A", "clinpgxlevel": "1A",
82	"pgxtesting": "Actionable PGx", "citations": ["21716271", ...]}
83	],
84	"recommendations": [
85	{"drugid": "RxNorm:32968",
86	"phenotypes": {"CYP2C19": "Ultrarapid Metabolizer"},
87	"implications": {"CYP2C19": "Increased active metabolite ..."},
88	"recommendation": "Use at standard dose (75 mg/day)",
89	"classification": "Strong",
90	"population": "CVI ACS PCI"}
91	]
92	}
93	]
94	```
95
96	On error: `{"query": "xxx", "error": "..."}`.
97
98	---
99
100	## Lower-level functions
101
102	\| Function \| Input \| Output \| Description \|
103	\|----------\|-------\|--------\|-------------\|
104	\| `get_drug_info(drug_name)` \| drug name \| `list[dict]` \| Drug table lookup (fuzzy) \|
105	\| `get_guidelines(drug_name=None)` \| optional drug name \| `list[dict]` \| All or filtered guidelines \|
106	\| `get_gene_drug_pairs(drug_name=None, gene=None)` \| optional filters \| `list[dict]` \| Gene-drug pairs (name auto-resolved to drugid) \|
107	\| `get_recommendations(drug_name)` \| drug name \| `list[dict]` \| Dosing recommendations (name auto-resolved) \|
108
109	---
110
111	## Notes
112
113	- CPIC levels: A = guideline published, B = in progress, C/D = lower evidence.
114	- Gene symbols are auto-detected (uppercase, ≤12 chars) and routed to `genesymbol` filter.
115	- Drug names are fuzzy-matched via `ilike` on the `/v1/drug` table.
116	- No rate limit documented, but keep requests reasonable.

People who liked this also viewed

6 picks

Type