Skip to content
RepoReady
Sign up
Menu
Docs menu

Materials: code, inputs and software

L1 Materials available needs three things for every key result: the code that computes it, the inputs it reads and the software it needs.

Last updated

One input is confirmed missing, so this project stays at L0 until it is available.
Level relationshipRequired for L1

Without the code, data and software behind a result, no documentation lets anyone check it, so Materials decide L1. Supporting results, such as supplementary figures, count one level down: their gaps block L2. The FAIR principles ask that data can be retrieved through a persistent identifier.

Code for each result is available

What counts
Public code, in a repository or archive, computes the values in every key figure, table and dataset. Statistics inside plotting scripts count; cosmetic plotting alone does not.
Common gaps
No code linked, a notebook never shared, code that computes a different quantity, or code last changed long before acceptance (a Warning).
How to fix
Share every script that computes a reported value, analyses added during revision included, and name it in the README.

Input data can be obtained

What counts
Every file the code reads is in the repository, produced by the code, or deposited with a working identifier. Controlled data count with a documented access route.
Common gaps
A file shared nowhere, a deposit lacking it, a dead link, or “available on request”. Data read by scripts behind no result: a Warning.
How to fix
Deposit data you produced in an archive that issues DOIs. For controlled data, name the committee, how to apply and where the files go:
README.md
## Data

| Dataset        | Source  | Identifier             | Version | Access                |
|----------------|---------|------------------------|---------|-----------------------|
| Raw counts     | Zenodo  | 10.5281/zenodo.0000000 | v2      | open                  |
| Patient cohort | EGA     | EGAD00000000000        | 2024-03 | data access committee |
| Reference      | Ensembl | GRCh38                 | 110     | open                  |

Required software can be obtained

What counts
Besides your own code, every package, tool, model and service it depends on, such as scanpy or a pretrained model, can be obtained from a package index, model hub, public repository or archive. Commercial software counts.
Common gaps
An in-house package never released, model weights not shared, a deleted repository without an archive. Steps that need commercial software: a Warning.
How to fix
Release in-house tools and weights, or archive the versions you used. Name commercial software with version, vendor and the steps that need it.

See where your project stands

Run the analysis on your paper or repository. Every finding names its area and links back here.

Get started