replicate-study

Name: replicate-study
Author: Aperivue

40installs53forks

How to install replicate-study?

aperivue/medsci-skills/replicate-study

$npx -y skills add Aperivue/medsci-skills --skill replicate-study

Installs into the current project.

Summary

Replicate an existing cohort study's methodology on a different database. Extracts study design from a source paper, maps variables to the target DB via harmonization table, generates analysis code, and produces a replication difference report.

Files · 1

View on GitHub

SKILL.md

1	# Replicate Study Skill
2
3	You are assisting a medical researcher in replicating an existing published study's methodology
4	on a different database. This is a common research strategy: take a validated methodology from
5	Paper A (e.g., NHIS cohort study) and apply it to Database B (e.g., KNHANES, NHANES, or another
6	cohort) to produce a new paper with the same analytical rigor.
7
8	## When to Use
9
10	- Researcher has a published paper they want to replicate on their own data
11	- Swapping exposure/outcome variables within the same DB
12	- Cross-national replication (e.g., Korean study → US data, or vice versa)
13	- Extending a single-institution study to a national cohort
14
15	## Inputs
16
17	1. Source paper: PDF, DOI, or markdown of the paper to replicate
18	2. Target database path: CSV/SAS data file(s) to use
19	3. Harmonization table (optional): CSV mapping source → target variables
20	- Default: `${SKILL_DIR}/references/harmonization_knhanes_nhanes.csv` (if KNHANES↔NHANES)
21
22	## Reference Files
23
24	- `${SKILL_DIR}/references/methodology_extraction_template.md` — checklist for extracting study design
25	- `${SKILL_DIR}/references/harmonization_knhanes_nhanes.csv` — KNHANES↔NHANES variable mapping (67 rows)
26	- `${SKILL_DIR}/references/harmonization_3country.csv` — KNHANES+NHANES+CHNS 3-country mapping (45 rows, if available)
27	- Upstream templates (read on demand):
28	- `medsci-skills/skills/write-paper/references/paper_types/nhis_cohort.md`
29	- `medsci-skills/skills/write-paper/references/paper_types/cross_national.md`
30	- `medsci-skills/skills/analyze-stats/references/analysis_guides/survey_weighted.md`
31	- `medsci-skills/skills/analyze-stats/references/analysis_guides/propensity_score.md`
32
33	## Workflow
34
35	### Phase 1: Source Paper Analysis
36
37	1. Read the source paper (PDF → text, or markdown).
38	2. Extract methodology using the extraction template:
39	- Study design: cohort / cross-sectional / case-control
40	- Database: name, country, years, N
41	- Population: inclusion/exclusion criteria, age range
42	- Exposure: variable name, definition, coding
43	- Outcome: variable name, definition, coding
44	- Covariates: full list with definitions
45	- Statistical methods: regression type, adjustment model, subgroup analyses
46	- Survey design: weights, strata, PSU (if applicable)
47	- Sensitivity analyses: list all
48	3. Output: structured extraction summary for user review.
49
50	### Phase 2: Variable Mapping
51
52	1. Load the harmonization table (CSV with columns: domain, concept, source_var, target_var, notes).
53	2. For each extracted variable (exposure, outcome, covariates):
54	- Find the matching row in the harmonization table
55	- Flag: DIRECT_MATCH / RECODE_NEEDED / NOT_AVAILABLE / PROXY_AVAILABLE
56	3. Generate a mapping report:
57	- Green: directly available (no recoding)
58	- Yellow: available but needs recoding (document transformation)
59	- Red: not available in target DB (propose proxy or exclusion)
60	4. Output: variable mapping table for user approval.
61
62	### Phase 3: Code Generation
63
64	1. Generate analysis code (Python with `pandas` + R via `subprocess` for survey-weighted):
65	a. Data loading & cleaning: read target DB, apply inclusion/exclusion
66	b. Variable derivation: recode variables per mapping table
67	c. Survey design setup: define svydesign object (strata, PSU, weights)
68	d. Table 1: demographics by exposure group (weighted)
69	e. Main analysis: replicate the primary model (logistic/Cox/linear regression)
70	f. Subgroup analyses: if specified in source paper
71	g. Sensitivity analyses: replicate all listed in source paper
72	2. Use `/analyze-stats` templates where available (survey_weighted, propensity_score).
73	3. All code must be self-contained and reproducible.
74
75	### Phase 4: Difference Report
76
77	Generate a structured difference report documenting:
78
79	\| Section \| Content \|
80	\|---------\|---------\|
81	\| Study Design \| Same / Modified (explain) \|
82	\| Database \| Source DB → Target DB (N, years, country) \|
83	\| Population \| Inclusion/exclusion differences \|
84	\| Variable Mapping \| Full mapping table with match status \|
85	\| Unavailable Variables \| What's missing and how handled \|
86	\| Methodological Differences \| Any forced changes (e.g., BMI cutoffs, LDL calculation) \|
87	\| Expected Differences \| Why results may differ (population, measurement, cultural) \|
88
89	Save as `replication_report.md` in the working directory.
90
91	### Phase 5: Validation Checklist
92
93	Before reporting completion, verify:
94
95	- [ ] All source paper covariates accounted for (mapped, proxied, or documen

People who liked this also viewed

6 picks

Type

replicate-study

40installs53forks

View on GitHub

How to install replicate-study?

aperivue/medsci-skills/replicate-study

$npx -y skills add Aperivue/medsci-skills --skill replicate-study

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Replicate Study Skill
2
3	You are assisting a medical researcher in replicating an existing published study's methodology
4	on a different database. This is a common research strategy: take a validated methodology from
5	Paper A (e.g., NHIS cohort study) and apply it to Database B (e.g., KNHANES, NHANES, or another
6	cohort) to produce a new paper with the same analytical rigor.
7
8	## When to Use
9
10	- Researcher has a published paper they want to replicate on their own data
11	- Swapping exposure/outcome variables within the same DB
12	- Cross-national replication (e.g., Korean study → US data, or vice versa)
13	- Extending a single-institution study to a national cohort
14
15	## Inputs
16
17	1. Source paper: PDF, DOI, or markdown of the paper to replicate
18	2. Target database path: CSV/SAS data file(s) to use
19	3. Harmonization table (optional): CSV mapping source → target variables
20	- Default: `${SKILL_DIR}/references/harmonization_knhanes_nhanes.csv` (if KNHANES↔NHANES)
21
22	## Reference Files
23
24	- `${SKILL_DIR}/references/methodology_extraction_template.md` — checklist for extracting study design
25	- `${SKILL_DIR}/references/harmonization_knhanes_nhanes.csv` — KNHANES↔NHANES variable mapping (67 rows)
26	- `${SKILL_DIR}/references/harmonization_3country.csv` — KNHANES+NHANES+CHNS 3-country mapping (45 rows, if available)
27	- Upstream templates (read on demand):
28	- `medsci-skills/skills/write-paper/references/paper_types/nhis_cohort.md`
29	- `medsci-skills/skills/write-paper/references/paper_types/cross_national.md`
30	- `medsci-skills/skills/analyze-stats/references/analysis_guides/survey_weighted.md`
31	- `medsci-skills/skills/analyze-stats/references/analysis_guides/propensity_score.md`
32
33	## Workflow
34
35	### Phase 1: Source Paper Analysis
36
37	1. Read the source paper (PDF → text, or markdown).
38	2. Extract methodology using the extraction template:
39	- Study design: cohort / cross-sectional / case-control
40	- Database: name, country, years, N
41	- Population: inclusion/exclusion criteria, age range
42	- Exposure: variable name, definition, coding
43	- Outcome: variable name, definition, coding
44	- Covariates: full list with definitions
45	- Statistical methods: regression type, adjustment model, subgroup analyses
46	- Survey design: weights, strata, PSU (if applicable)
47	- Sensitivity analyses: list all
48	3. Output: structured extraction summary for user review.
49
50	### Phase 2: Variable Mapping
51
52	1. Load the harmonization table (CSV with columns: domain, concept, source_var, target_var, notes).
53	2. For each extracted variable (exposure, outcome, covariates):
54	- Find the matching row in the harmonization table
55	- Flag: DIRECT_MATCH / RECODE_NEEDED / NOT_AVAILABLE / PROXY_AVAILABLE
56	3. Generate a mapping report:
57	- Green: directly available (no recoding)
58	- Yellow: available but needs recoding (document transformation)
59	- Red: not available in target DB (propose proxy or exclusion)
60	4. Output: variable mapping table for user approval.
61
62	### Phase 3: Code Generation
63
64	1. Generate analysis code (Python with `pandas` + R via `subprocess` for survey-weighted):
65	a. Data loading & cleaning: read target DB, apply inclusion/exclusion
66	b. Variable derivation: recode variables per mapping table
67	c. Survey design setup: define svydesign object (strata, PSU, weights)
68	d. Table 1: demographics by exposure group (weighted)
69	e. Main analysis: replicate the primary model (logistic/Cox/linear regression)
70	f. Subgroup analyses: if specified in source paper
71	g. Sensitivity analyses: replicate all listed in source paper
72	2. Use `/analyze-stats` templates where available (survey_weighted, propensity_score).
73	3. All code must be self-contained and reproducible.
74
75	### Phase 4: Difference Report
76
77	Generate a structured difference report documenting:
78
79	\| Section \| Content \|
80	\|---------\|---------\|
81	\| Study Design \| Same / Modified (explain) \|
82	\| Database \| Source DB → Target DB (N, years, country) \|
83	\| Population \| Inclusion/exclusion differences \|
84	\| Variable Mapping \| Full mapping table with match status \|
85	\| Unavailable Variables \| What's missing and how handled \|
86	\| Methodological Differences \| Any forced changes (e.g., BMI cutoffs, LDL calculation) \|
87	\| Expected Differences \| Why results may differ (population, measurement, cultural) \|
88
89	Save as `replication_report.md` in the working directory.
90
91	### Phase 5: Validation Checklist
92
93	Before reporting completion, verify:
94
95	- [ ] All source paper covariates accounted for (mapped, proxied, or documen

People who liked this also viewed

6 picks

Type