D-QUARK results for QD1

Input Sequence in FASTA format

>protein (51 residues)
VVYTDCTESGQNLCLCEGSNVCGQGNKCILGSDGEKNQCVTGEGTPKPQSH

Predicted Secondary Structure

Sequence	20 40 \| \| VVYTDCTESGQNLCLCEGSNVCGQGNKCILGSDGEKNQCVTGEGTPKPQSH
Prediction	CCCSCCCCCCCCSSSSCCCCCCCCCCCSSSCCCCCCCCSSCCCCCCCCCCC
Confidence	940405755660599715774677990584899985531336786699999
	H:Helix; S:Strand; C:Coil

Predicted distance map

o d_ij<8Å o d_ij<20Å

D-QUARK simulation is guided by distance-map predicted by DeepPotential for Cα-Cα and Cβ-Cβ interactions, shown in the lower right and upper left triangles in the above image, respectively. Dark and light circles are for distances predicted to be within 8 and 20 Å, respectively.

Top 5 final models

Rank	Download	Estimated TM-score	Estimated RMSD
1	model1.pdb.gz	0.4766	9.35 Å
2	model2.pdb.gz	0.4748	9.89 Å
3	model3.pdb.gz	0.4726	10.02 Å
4	model4.pdb.gz	0.4711	10.20 Å
5	model5.pdb.gz	0.3788	10.96 Å

Read more about estimated TM-score and estimated RMSD here

References:
Chengxin Zhang, Yang Li, Yang Zhang (2021) D-QUARK: ab initio protein structure prediction guided by multiple deep learning predicted distance and orientation restraints.