The user guide of RNA Splicer


Mutations in exon or even in intron may lead to changes in splicing recognition sequences or other regulatory factors affecting exons. splicing, eventually leading to more serious consequences. According to the summary of the human SNP pathogenic data on the malacards website (, it can be found that no less than 15% of the pathogenic mutations in the data are in introns, and most of these intronic pathogenic mutations occur near the exon and intron boundaries. These mutations mostly affect mRNA splicing.
RNA splicer is an AI model tools developed by the Artificial Intelligence Innovation Center of Tsinghua Pearl River Delta Research Institute. It predicts the alternative splicing sites caused by base mutations. The AI model of RNA Splicer was implemented by the machine learning HMM method combined with the permutation test. It predicts the alternative splicing site of the mRNA sequence by learning the canonical splicing pattern.
Existing splicing prediction tools generally focus on predicting the effect of base mutations on normal splicing positions and alternative splicing sites. However, no specific analysis has been carried out on the mechanism leading to the alternative splicing site and the resulting consequences. At the same time, the existing prediction tools are difficult to use and interpret the results. Some tools even require users to build their own algorithm models to convert the data, and there is no visual display of the results. On the other hand, RNA Splicer has the friendly and interactive interface, data acquisition is simple, results are easy to interpret, and the mechanism of generating alternative splicing sites and the results are displayed. The results are relatively informative.

Instruction Manual

Prediction information

The length of SNV should be less than 50
For mutation information, it provides three sequence input methods: "g." indicates a DNA reference sequence,"c." indicates a coding DNA reference sequence,"p." indicates a protein reference sequence

"g." indicates a DNA reference sequence

Examples: g.95T>G, a substitution of the T nucleotide at g.95 (DNA reference sequence) by a G

"c." indicates a coding DNA reference sequence

Examples: c.65T>G, a substitution of the T nucleotide at c.65 (coding DNA reference sequence) by a G
Examples: c.68+1G>A, a substitution of the T nucleotide at c.68+1 (coding DNA reference sequence) by a G
Examples: c.52_53delTG, a deletion of nucleotides c.52 to c.53 (coding DNA reference sequence)
Examples: c.52_53insAGG, the insertion of nucleotides AGG between nucleotides c.52 and c.53 (coding DNA reference sequence)
Examples: c.52_53delTGinsAGG, a deletion replacing nucleotides c.52 to c.53 (TG, not described) with AGG
Examples: c.-12C>T, nucleotides C altered to T at the last 12th position in 5`UTR
Examples: c.*17G>T, nucleotides G altered to T at the 17th position in 3`UTR

"p."indicates a protein reference sequence

Examples: p.A3F(GCT>TTT), a substitution of the Alanine at p.3 (protein reference sequence) by Phenylalanine

View Results

The results page will display the molecular results in the upper left corner of the splicing map:
Indicates the length of bases inserted or deleted
Whether the sequence change causes frameshift of the protein coding frame, whether the sequence coding is terminated early
Whether the inserted sequence exists in the form of a pseudo-exon.


Original splicing site
New splicing site
Psuedo exon
Other potential backup splicing site
Predicted score decreased after mutation / increase
Exon concatenation
The predicted score decreased after mutation

Results display

The predicting RNA splicing results may include the following situations:
Original splicing pattern: consistent with the wild-type splicing pattern, if the score after mutation is significantly lower than that before mutation, the expression of RNA may be affected by base mutation
The new identified splicing site is located in exon, the donor or receptor was replaced by the new possible splicing site so that the part of exon was clipped.
The original splicing site sequence was mutated, meanwhile the new possible splicing site may be located in the neighbor intron.
The original splicing site sequence was mutated so that the potential backup splicing site was activated.
The emerging pseudo exon was caused by the mutations far from the edge of the intron.
The exon skipping was caused by the mutation in the original splicing site, and the backup splicing site was lack in the neighbor intron.


Al agent
Back to top