About

1 Variant prediction

1.1 Input Instructions

PDB file can be downloaded by providing ID. Also the chain ID has to be provided. (e.g. >1aj3-A)

Provide the variants in HGVS format, i.e. original amino acid, position number, variant amino acid (e.g. P61A).

For prediction type, users can choose between regression and classification prediction.

---

1.2 Output

2 Protein prediction

Predict all possible 19 single amino acid substitutions in all positions in a single protein. Provide PDB and chain id.

2.1 Input Instructions

PDB file can be downloaded by providing ID. Also the chain ID has to be provided. (e.g. >2jof-A)

For prediction type, users can choose between regression and classification prediction.

It will take a long time, please provide your email and be patient.

2.2 Output

Prediction result for all position of the protein.  example

Data sets

Extensive data mining was performed for obtaining cases for training and testing.  download

Dataset	variations	decreasing	no effect	increasing
Training dataset	762	520	136	106
Blind Test dataset	190	133	39	18

The source paper of the data sets.

PDB ID	Number of variants	PubMed ID
1AJ3	2	15504411
1APS	24	10542090
1BNR	19	15035606
1FKB	34	10438631
1HMK	17	15276846
1IMQ	32	12547210
1N88	17	15136744
1SCE	56	10673431
1TEN	39	10704314
1TIT	26	11377196
2ABD	30	10360367
2CI2	68	7490748
2PTL	68	10801362
3GB1	31	10932252
1CSP	22	15147842
1DIV-C	24	17240985
1DIV-N	14	16246369
1E0L	45	16784750
1FMK	57	10542092
1O6X	18	9799641
1RFA	47	17137592
1SHF	34	11786916
1SHG	14	10542091
1SS1	63	17628591
1SSO	23	11124040
1ST7	18	15690348
1UBQ	27	15857839
1W4E	22	16406408
1W4J	22	18625240
1YYJ	39	12069590

Contact

If you have any problems, please contact Chong Zhang (20205227090@stu.suda.edu.cn).