SPIDEX

Our first product is SPIDEXTM, a comprehensive set of genetic variants and their predicted effects on human splicing across the entire genome. The most advanced, highest quality version of SPIDEXTM is available for R&D licensing and for per sample clinical analysis. We are making the SPIDEXTM database free for non-commercial purposes.

SPIDEX TM provides predictions of how 328 million single nucleotide variants (SNVs) may affect RNA splicing in humans. It addresses a key unmet challenge in genomic medicine, which is to ascertain how SNVs cause splicing misregulation and may lead to disease. Unlike other approaches, SPIDEX TM is produced by a direct statistical model of the splicing mechanism whose predictive power for variants is independent of their population frequencies. As a result, it can be used to analyze rare and de novo mutations. SPIDEX TM covers all synonymous, missense and nonsense exonic SNVs, as well as intronic SNVs that are up to 300nt from splice junctions.

SPIDEX TM is produced by our state-of-the-art splicing simulator assembled by our proprietary deep learning algorithms. It is a Bayesian ensemble of deep neural networks trained with RNA sequencing data from a diverse set of healthy human tissues and thousands of carefully engineered RNA features. It captures 65% of variance of splicing levels across exon triplets observed to undergo alternative cassette splicing (read the paper in Science). When tested on genetic variants that underlie cystic fibrosis, spinal muscular atrophy and hereditary nonpolyposis colorectal cancer, it achieves unprecedented predictive performance, especially for deleterious mutations ignored by competing technologies, such as intronic mutations that are far from splice sites.

 
 

Out-of-sample performance of SPIDEX TM for documented splicing related mutations found in four disease causing genes. Mutations included in the above plot are either known to be deleterious or benign or known to enhance or decrease splicing.

How to get SPIDEX

We are making SPIDEX TM available for free to non-commercial users for research purposes. To inquire about licensing SPIDEX TM for commercial use, please contact us.