Dataset Viewer
Auto-converted to Parquet Duplicate
target_id
stringlengths
24
27
sequence
stringlengths
27
258
length
int64
27
258
pdb_id
stringlengths
4
4
auth_chain_id
stringlengths
1
4
source_split
stringclasses
1 value
rna3db_component
stringclasses
33 values
rna3db_cluster_id
stringlengths
6
9
conformer_index
int64
0
0
structure_hash
stringlengths
64
64
native_ceiling_gdt_ts
float64
0.3
0.96
native_ceiling_tm_score
float64
0.13
0.97
native_ceiling_rmsd
float64
0.71
9.55
ref_c4p_coords
listlengths
81
774
ref_backbone_coords
listlengths
243
2.32k
mask_coords
listlengths
27
258
rna3db_4p8z_A_f38ef6ae5d
GGUUGGGUUGGGAAGUAUCAUGGCUAAUCACCAUGAUGCAAUCGGGUUGAACACUUAAUUGGGUUAAAACGGUGGGGGACGAUCCCGUAACAUCCGUCCUAACGGCGACAGACUGCACGGCCCUGCCUCUUAGGUGUGUCCAAUGAACAGUCGUUCCGAAAGGAAGCAUCCGGUAUCCCAAGACAAUC
188
4P8Z
A
rna3db_train
component_22
4p8z_A
0
d2dd2f3bcbbe78df71a88c1bf0b8a1c3b7b6cc5091844ba64692e3d0d20c83c9
0.3045
0.5255
5.732
[ 41.064998626708984, -14.267999649047852, -52.96200180053711, 43.6349983215332, -12.057999610900879, -47.95000076293945, 44.14500045776367, -7.827000141143799, -43.46500015258789, 41.0260009765625, -3.2130000591278076, -40.87900161743164, 35.82500076293945, -0.19699999690055847, -40.14400...
[ 37.292999267578125, -13.755000114440918, -53.939998626708984, 41.064998626708984, -14.267999649047852, -52.96200180053711, 42, -11.088000297546387, -53.53499984741211, 40.28300094604492, -13.682000160217285, -49.18299865722656, 43.6349983215332, -12.057999610900879, -47.95000076293945, ...
[ true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true...
rna3db_6ft6_2_9b1ce89a27
AAACUUUCAACAACGGAUCUCUUGGUUCUCGCAUCGAUGAAGAACGCAGCGAAAUGCGAUACGUAAUGUGAAUUGCAGAAUUCCGUGAAUCAUCGAAUCUUUGAACGCACAUUGCGCCCCUUGGUAUUCCAGGGGGCAUGCCUGUUUGAGCGUCAUAAAAUU
162
6FT6
2
rna3db_train
component_1
6ft6_2
0
d44e023531f781bddfd7aec878d2d941e3485faa72bbdba1f35192d86bb1bbdb
0.3086
0.4824
7.273
[ 355.9219970703125, 295.65399169921875, 234.23199462890625, 356.3089904785156, 296.6570129394531, 240.49200439453125, 353.64300537109375, 298.90899658203125, 245.43299865722656, 348.6080017089844, 297.8160095214844, 248.30299377441406, 344.6180114746094, 293.5159912109375, 249.69099426269...
[ 352.61700439453125, 293.9429931640625, 233.45700073242188, 355.9219970703125, 295.65399169921875, 234.23199462890625, 354.06500244140625, 298.1759948730469, 235.3249969482422, 356.3999938964844, 294.0069885253906, 237.73500061035156, 356.3089904785156, 296.6570129394531, 240.492004394531...
[ true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true...
rna3db_8q5i_4_bd654503ba
AAACUUUCAACAACGGAUCUCUUGGUUCUCGCAUCGAUGAAGAACGCAGCGAAAUGCGAUACGUAAUAUGAAUUGCAGAUAUUCGUGAAUCAUCGAAUCUUUGAACGCACAUUGCGCCCUCUGGUAUUCCGGAGGGCAUGCCUGUUUGAGCGUCGUUU
158
8Q5I
4
rna3db_train
component_1
9g30_4
0
bbd94fdd48b53bac490f08a29104973dfaf9d5e4090a61f2218e8de8860397d9
0.3006
0.4672
6.589
[ 166.73300170898438, 251.13600158691406, 148.78799438476562, 171.83700561523438, 254.48500061035156, 151.06199645996094, 175.93699645996094, 255.26499938964844, 155.17100524902344, 179.6300048828125, 251.3350067138672, 157.67799377441406, 182.6060028076172, 246.04200744628906, 156.6600036...
[ 167.63099670410156, 247.45599365234375, 148.6479949951172, 166.73300170898438, 251.13600158691406, 148.78799438476562, 166.85400390625, 251.2480010986328, 152.1060028076172, 170.60899353027344, 251.86199951171875, 148.50999450683594, 171.83700561523438, 254.48500061035156, 151.0619964599...
[ true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true...
rna3db_7u0h_2_00f41a5246
AAACUUUCAACAACGGAUCUCUUGGUUCUCGCAUCGAUGAAGAACGCAGCGAAAUGCGAUACGUAAUGUGAAUUGCAGAAUUCCGUGAAUCAUCGAAUCUUUGAACGCACAUUGCGCCCCUUGGUAUUCCAGGGGGCAUGCCUGUUUGAGCGUCAUUU
158
7U0H
2
rna3db_train
component_1
7osa_58S
0
a3fb58f76ba691ad41ec08fa1f7fdcf98680df5eb09c190e514db5f2d1d7d4ab
0.3449
0.5336
5.727
[ 242.98904418945312, 233.62704467773438, 249.0445556640625, 244.7400360107422, 231.59603881835938, 243.07354736328125, 244.08204650878906, 227.8680419921875, 238.383544921875, 239.14503479003906, 225.86703491210938, 235.60154724121094, 233.0330352783203, 226.8910369873047, 234.28555297851...
[ 239.43203735351562, 233.05303955078125, 250.3385467529297, 242.98904418945312, 233.62704467773438, 249.0445556640625, 243.123046875, 230.28004455566406, 248.77354431152344, 243.10704040527344, 234.4380340576172, 245.20755004882812, 244.7400360107422, 231.59603881835938, 243.0735473632812...
[ true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true...
rna3db_7r81_C1_22fa693bb4
AAACUUUCAACAACGGAUCUCUUGGUUCUGGCAUCGAUGAAGAACGCAGCGAAAUGCGAUAGGUAAUGUGAAUUGCAGAAUUCAGUGAAUCAUCGAAUCUUUGAACGCACAUUGCGCUCGCCAGUAUUCUGGCGAGCAUGCCUGUUCGAGCGUCAUUU
158
7R81
C1
rna3db_train
component_1
7r81_C1
0
c3f9535c3d96024d1ebb7fbd34babd4d39509e671a28e1faebc83e65182b842d
0.3987
0.5776
5.094
[ 255.01499938964844, 240.41299438476562, 265.135986328125, 250.28700256347656, 243.71200561523438, 262.8689880371094, 245.77000427246094, 243.7429962158203, 258.7309875488281, 241.84100341796875, 239.78700256347656, 256.3299865722656, 238.95899963378906, 234.4010009765625, 257.54901123046...
[ 254.8820037841797, 236.51699829101562, 265.05999755859375, 255.01499938964844, 240.41299438476562, 265.135986328125, 255.30299377441406, 240.39199829101562, 261.8340148925781, 251.2100067138672, 240.75900268554688, 265.260009765625, 250.28700256347656, 243.71200561523438, 262.86898803710...
[ true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true, true...
rna3db_5it7_8_c27a83d905
"AAACUUUCAACAACGGAUCUCUUGGUUCUCGCAUCGAUGAAGAACGCAGCGAAUUGCGAUAUGUAUUGUGAAUUGCAGAUUUUCGUGAAUCAUCAAAUC(...TRUNCATED)
157
5IT7
8
rna3db_train
component_1
6uz7_8
0
0a837edc517596d25b2271c7c82559417f25a0db4f6d45f202348df68a306c0f
0.3678
0.5298
6.418
[299.4200134277344,194.31700134277344,189.46400451660156,299.3489990234375,196.01300048828125,195.90(...TRUNCATED)
[296.22698974609375,192.24200439453125,189.3179931640625,299.4200134277344,194.31700134277344,189.46(...TRUNCATED)
[true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true(...TRUNCATED)
rna3db_9axu_4_23a98f2aa8
"AAACUUUCAGCAACGGAUCUCUUGGCUCUCGCAUCGAUGAAGAACGCAGCGAAAUGCGAUACGUAAUGUGAAUUGCAGAAUUCCGUGAAUCAUCGAAUC(...TRUNCATED)
157
9AXU
4
rna3db_train
component_1
8ev3_2
0
442725872c43a46d0ca355028bc3c9ae6529739982f1946dd65bf57153cd910c
0.3392
0.5251
5.867
[158.28305053710938,180.58758544921875,268.6856994628906,163.1588134765625,177.6460723876953,265.896(...TRUNCATED)
[159.14230346679688,184.27821350097656,269.7688293457031,158.28305053710938,180.58758544921875,268.6(...TRUNCATED)
[true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true(...TRUNCATED)
rna3db_9ndp_8_dc2bbdbc91
"GACUCUUAGCGGUGGAUCACUCGGCUCGUGCGUCGAUGAAGAACGCAGCUAGCUGCGAGAAUUAAUGUGAAUUGCAGGUUGAUCAUCGACACUUCGAAC(...TRUNCATED)
150
9NDP
8
rna3db_train
component_1
6d9j_8
0
6659434a64b8b5d97c9f2da206fa93c1f8d9937cda85dc7f33a1d5bd80053cce
0.345
0.5247
5.749
[156.30299377441406,236.70799255371094,170.67300415039062,160.06100463867188,237.93299865722656,175.(...TRUNCATED)
[157.7010040283203,235.9459991455078,167.0780029296875,156.30299377441406,236.70799255371094,170.673(...TRUNCATED)
[true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true(...TRUNCATED)
rna3db_5m73_A_edd8c12495
"GGGUGUCCGCACUAAGUUCGGCAUCAAUAUGGUGACCUCCCGGGAGCGGGGGACCACCAGGUUGCCUAAGGAGGGGUGAACCGGCCCAGGUCGGAAACG(...TRUNCATED)
145
5M73
A
rna3db_train
component_1
7nfx_1
0
ea217e24081eb227e04a307802f709f86e71b95dd59c1e58565c915fb817b099
0.5155
0.696
3.182
[-25.863000869750977,84.24099731445312,-16.98200035095215,-24.827999114990234,78.8949966430664,-14.2(...TRUNCATED)
[-27.457000732421875,86.24800109863281,-19.481000900268555,-25.863000869750977,84.24099731445312,-16(...TRUNCATED)
[true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true(...TRUNCATED)
rna3db_6n5q_A_4e73325013
"GGUUGCCGAAUCCACCUAGAAAUGGUACGGAGGAACCGCUUUUUGGGGUUAAUCUGCAGUGAAGCUGCAGUAGGGAUACCUUCUGUCCCGCACCCGACA(...TRUNCATED)
128
6N5Q
A
rna3db_train
component_1
6n5q_A
0
21a40f423000742a51c3057f24dc7c1560aa80561aff83a409563f8e83fca665
0.4707
0.6198
5.246
[-61.72200012207031,5.2829999923706055,-27.104000091552734,-65.16899871826172,9.444000244140625,-29.(...TRUNCATED)
[-59.62799835205078,3.2019999027252197,-29.600000381469727,-61.72200012207031,5.2829999923706055,-27(...TRUNCATED)
[true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true(...TRUNCATED)
End of preview. Expand in Data Studio

RIDE-RL RNA Inverse Folding Pools

Curated RNA target pools for reinforcement-learning post-training of RNA inverse-folding models, and for evaluating them. 527 training targets and 153 held-out test targets, drawn from RNA3DB and verified to be single, independently-foldable chains.

Companion code: https://github.com/Gabrile166/RIDE-RL

Why these pools exist

Two problems made the obvious choices unusable.

Composites do not fold on their own. Selecting chains by resolution, date and length alone yields many that are one piece of a larger assembly: one strand of a duplex, a segment threaded through a protein, a fragment of a ribosome. The PDB entry records the conformation the chain holds inside that assembly. Fold its sequence alone and you get the conformation it adopts alone, which is a different thing. In a batch selected this way, the native sequences themselves folded back to a median GDT-TS of 0.053. When the native sequence cannot recover its own reference, no designed sequence can score on that target either, so every model collapses to noise and the comparison measures nothing.

A pool must be learnable to be informative. Targets that are trivially easy or hopelessly hard both carry little signal.

Both pools are therefore filtered on a foldability criterion: the native sequence, folded by RhoFold+ and compared against the deposited structure, must reach GDT-TS >= 0.30. Median native ceiling is 0.64 (train) and 0.60 (test) -- see native_ceiling_* below.

Splits

split targets length range median length median native ceiling (GDT-TS)
train 527 27-258 nt 75 0.640
test 153 15-186 nt 65 0.601

Disjointness is verified, not assumed: the two splits share zero sequences and zero target ids.

Fields

field description
target_id stable identifier, rna3db_<pdb>_<chain>_<hash>
sequence native RNA sequence (ACGU)
length sequence length in nucleotides
pdb_id, auth_chain_id provenance in the PDB
source_split originating RNA3DB split
rna3db_component, rna3db_cluster_id RNA3DB structural clustering, for diversity accounting
conformer_index distinguishes multiple conformers of one chain
structure_hash hash of the reference coordinates
native_ceiling_gdt_ts GDT-TS of the native sequence folded back by RhoFold+
native_ceiling_tm_score same, TM-score
native_ceiling_rmsd same, RMSD in angstrom
ref_c4p_coords reference C4' coordinates, flattened, length x 3
ref_backbone_coords reference backbone coordinates, flattened, length x 3 x 3
mask_coords per-residue validity mask, length length

Reading the coordinates

Both coordinate fields are flattened for portability. Restore them with:

import json
import numpy as np

rec = json.loads(open("test.jsonl").readline())
n = rec["length"]
c4p = np.array(rec["ref_c4p_coords"]).reshape(n, 3)
backbone = np.array(rec["ref_backbone_coords"]).reshape(n, 3, 3)
mask = np.array(rec["mask_coords"], dtype=bool)

The native ceiling, and how to use it

native_ceiling_gdt_ts is the score obtained by folding the native sequence and comparing it to the deposited structure. It is the practical upper bound for that target under this evaluation protocol: a designed sequence is measured against the same reference through the same folding model, so it inherits the same error floor.

Report designs relative to it. Absolute scores conflate design quality with how well the folding model handles the target. Stratifying by ceiling also matters: in our own study, a reward that appeared to improve on the pool average turned out to be improving only on low-ceiling targets, where the reference is least trustworthy, while regressing on the high-ceiling ones. Averages hide that.

Provenance

Derived from RNA3DB. Reference coordinates come from RCSB PDB entries. Native ceilings were computed with RhoFold+, structural comparison with US-align. Construction scripts are under scripts/dataset/ in the companion repository; the pipeline and every filter threshold are documented in docs/DATASET_PIPELINE.md.

Limitations

  • Foldability screening uses RhoFold+, so the pools carry that model's inductive biases. A target excluded for a low native ceiling is not necessarily a poor structure; it may be one RhoFold+ handles badly.
  • RNA3DB component labels are coarse. component_1 dominates by label, though sequence-level diversity is high (median pairwise identity 0.27).
  • Multiple conformers of the same chain appear as separate records, distinguished by conformer_index. Group by pdb_id if that is not wanted.
  • No experimental secondary structures. Work needing them must derive them, e.g. by folding the native sequence, which is a prediction and not ground truth.

Citation

Manuscript in preparation. Please cite the repository until it appears.

Downloads last month
17