# RNA

## Inverse folding problem

The goal of the *inverse RNA folding problem* is to generate RNA sequences which fold into a pre-determined shape.

**Note:** The inverse RNA folding problem may be more of a sampling problem than a constraint satisfaction problem, since the number of constraints per ribonucleotide is much smaller but the number of off-target foldings is much higher.

### Datasets

- RNA STRAND v2.0 - Contains ~5000 RNA sequences and base-pairings. Has not been updated in ~10 years.
- bpRNA-1m - Contains ~100,000 RNA sequences and base-pairings. Published in 2018.
- Eterna100 - A representative dataset of RNA structures solved by Eterna players.
- RNAiFold 2.0 - 63 experimentally-synthesized targets.
- CASP RNA Challenge — CASP is planning to include RNA puzzles in the future.

### Software

- Vienna
- RNAStructure - Contains useful utilities, including
`dot2ct`

and`ct2dot`

, which allow the conversion between the CT file format and the Dot Bracket file format. - EternaBrain - Source code for
`EternaBrain-SAP`

. - CDPfold - Solve the inverse RNA folding problem using reinforcement learning.
- jadeshi/SentRNA - Solve the inverse RNA folding problem using reinforcement learning.