Predict allosteric site-forming residues from a protein structure

Give STINGAllo a PDB identifier and a chain, and it returns the residues it predicts form an allosteric site — at single-residue resolution, with an interactive 3D view of the result.

How it works

Free and open to all. No registration.

At a glance

Input
A 4-character PDB code and one chain, or your own .pdb file.
Output
The predicted allosteric site-forming residues (AFRs), rendered on the structure and downloadable.
Who it is for
Structural biologists, computational chemists and drug-discovery groups looking for allosteric sites.
Programmatic access
REST API for pipeline integration.

A per-residue approach, not a pocket search

In predicting allosteric sites, STINGAllo highlights the pivotal role of the internal protein nanoenvironment in defining the allosteric potential of residues. Traditional methods often focus on identifying pockets, which can miss crucial spatial and physicochemical contexts. StingAllo, however, employs a per-residue approach, in analyzing the nanoenvironment surrounding each residue, ensuring a more accurate prediction of allosteric forming residues (AFRs), surpassing the capabilities of pocket-focused techniques.

Nanoenvironment descriptors

Optimised internal descriptors per residue, including hydrophobic interaction networks, local density, graph connectivity and the “sponge effect” metric.

78%

Benchmark success rate

Against 21.1–24.2% for contemporary pocket-based predictors, which miss 18% of confirmed sites lying outside surface invaginations.

52.7%

Of unique PDB proteins

Contain at least one chain with a predicted allosteric site, across the 119,851 entries surveyed on 14 November 2024.

Read the STINGAllo publications

Getting started

Three steps, from a PDB code to a highlighted structure.

Enter a PDB code on the Server page
Type a 4-character code such as 3I54, or upload your own .pdb file. Worked examples are offered on the form.
Choose one chain
STINGAllo lists the chains in the entry, with organism, molecule and sequence length pulled from RCSB, so you can pick the right one. Exactly one chain per run.
Read the result
The predicted AFRs are listed as residue name and number, and drawn as a surface on the chain. You can export the structure, a PyMOL script and the residue list together.

Go to the STINGAllo Server

Global usage

Where STINGAllo has been accessed from.

Loading globe…

11467

Total visits worldwide

Countries reached

11467

Total visits

Visits by country, highest first
#	Country	Code	Visits
1	United States	US	6161
2	Brazil	BR	889
3	United Kingdom	GB	595
4	China	CN	467
5	Spain	ES	256
6	Netherlands	NL	193
7	Singapore	SG	189
8	Czech Republic	CZ	169
9	Vietnam	VN	158
10	India	IN	131
11	Germany	DE	112
12	Ireland	IE	108
13	France	FR	83
14	Hong Kong	HK	81
15	Moldova	MD	74
16	Bulgaria	BG	57
17	Mexico	MX	42
18	Japan	JP	37
19	Switzerland	CH	36
20	Canada	CA	35
21	Australia	AU	28
22	South Korea	KR	28
23	Iran	IR	24
24	Taiwan	TW	24
25	Indonesia	ID	23
26	Turkey	TR	20
27	Saudi Arabia	SA	17
28	Sweden	SE	17
29	Bangladesh	BD	15
30	Georgia	GE	15
31	Poland	PL	15
32	Lithuania	LT	12
33	Romania	RO	11
34	Pakistan	PK	9
35	Israel	IL	8
36	Italy	IT	8
37	Ukraine	UA	8
38	Denmark	DK	7
39	Algeria	DZ	6
40	Russia	RU	6
41	South Africa	ZA	6
42	Belgium	BE	5
43	Colombia	CO	5
44	Iraq	IQ	5
45	Nigeria	NG	5
46	New Zealand	NZ	5
47	Qatar	QA	5
48	Seychelles	SC	5
49	Argentina	AR	4
50	Finland	FI	4
51	United Arab Emirates	AE	3
52	Austria	AT	3
53	Egypt	EG	3
54	Chile	CL	2
55	Morocco	MA	2
56	Philippines	PH	2
57	Thailand	TH	2
58	Uruguay	UY	2
59	Bosnia and Herzegovina	BA	1
60	Bolivia	BO	1
61	Belarus	BY	1
62	Costa Rica	CR	1
63	Cyprus	CY	1
64	Ecuador	EC	1
65	Jamaica	JM	1
66	Jordan	JO	1
67	Cambodia	KH	1
68	Kuwait	KW	1
69	Latvia	LV	1
70	Mauritius	MU	1
71	Malaysia	MY	1
72	Norway	NO	1
73	Paraguay	PY	1
74	Serbia	RS	1
75	Slovakia	SK	1
76	Syria	SY	1
77	Togo	TG	1
78	Uzbekistan	UZ	1
79	Venezuela	VE	1
80	Nepal	NP	1