Table 2 Select list of in silico tools for predicting mRNA splicing effects
From: In silico methods for predicting functional synonymous variants
Tool | Algorithm/prediction method | Input | Output | URL/comments | Ref |
|---|---|---|---|---|---|
Splice site prediction tools | |||||
GENSCAN | Motif-based, maximal dependence decomposition (MDD) | Sequences up to 1 million nucleotides can be analyzed | Predicted exons and/or peptides in the sequence | http://hollywood.mit.edu/GENSCAN.html A copy of the program is available on request | [106] |
MaxEntScan | Motif-based, based on maximum entropy principle (MEP) | 9 and 23 nucleotide long sequences for donor and acceptor site predictions respectively | Tool provides a score for the sequence indicating its strength as splice site | http://hollywood.mit.edu/burgelab/maxent/Xmaxentscan_scoreseq.html Offers multiple scoring models as options Perl scripts to run algorithm are available for download | [107] |
Genesplicer | Motif-based, MDD with Markov models | Sequences of up to 200,000 nucleotides | Predicted acceptor and donor sites in the sequence with scores | https://www.cbcb.umd.edu/software/GeneSplicer/gene_spl.shtml Program is available for download | [108] |
NetGene2 | Machine-learning (ML) based, neural networks | One sequence between 200 and 100,000 nucleotides | Program provides predicted acceptor and donor sites with a confidence score | https://services.healthtech.dtu.dk/service.php?NetGene2-2.42 Program is available for download. The training dataset curated from NCBI GenBank included 65 human genes with 331 donor and acceptor splice sites | [109] |
Spliceport | ML-based, support vector machine | Sequence of up to 30,000 nucleotides | Spliceport provides a list a predictions for donor and acceptor sites along with a score | https://spliceport.cbcb.umd.edu/SplicingAnalyser2.html Training dataset included a collection of 4000 pre-mRNA human RefSeq sequences | [110] |
MMSplice | ML(deep learning)-based, modular neural networks | Predictive analysis of variants is performed on any exon with 50 and 13 nucleotides upstream and downstream respectively | Predicts effects of variants on exon skipping, splice site choice, splicing efficiency, and pathogenicity | Models are available in the Kipoi repository Trained on distinct large-scale genomics datasets. Please refer to Table 1 of citation for a detailed summary of trained modules and models | [111] |
SRE prediction tools | |||||
ESEfinder | Functional SELEX, PWM | Sequences of up to 5000 nucleotides | Tool provides predicted splice sites and SREs along with a score | http://krainer01.cshl.edu/cgi-bin/tools/ESE3/esefinder.cgi?process=home | [112] |
EX-SKIP | Predictions are based on a ratio of ESE/ESSs identified from 5 different models including RECUE-ESE, FAS-ESSs | Two exonic sequences strictly in uppercase in FASTA format up to a total length of 4000 nucleotides | Program compares the ESE/ESS profile of a native and variant sequence and predicts the probability of exon skipping | [113] | |
FAS-ESS | Experimental verification of random decanucleotide (10-nucleotides) sequences and weight matrices of aligned sequences | Single or multiple sequences in FASTA format. Length limit not specified | Program shows predicted ESS motifs. Users can search for hex2 or hex3 sets with relatively higher sensitivity and specificity respectively | [114] | |
ESRSeq | Experimental assessment of the SRE properties of all possible hexamer motifs | 6-hexamer sequences overlapping the variants | Calculate the net ESRseq score. Net score change could indicate prediction direction | Scores of ESE/ESS hexamers are available as supplementary materials | [115] |
Combination analysis tools | |||||
SROOGLE | Predictions based on 9 different models including MEP, PWM, ESEFinder, RESCUE-ESE, FAS-ESS, etc | Target exon along with flanking introns | Provides predictions for both splice sites and SREs | [116] | |
ExonScan | Splice site predictions based on MEP and SRE predictions based on RESCUE-ESE and FAS-ESS | DNA sequence with exons with at least 20 bases upstream of the exon it can predict and 60 bases of downstream of the last | Provides predictions for both splice sites and SREs | [114] | |