Quick Start Guide: predict

This guide provides a brief walkthrough for using the predict subcommand in OpenSpliceAI to predict splice sites from DNA sequences. If you haven't done so already, please see the Installation page for details on installing and configuring OpenSpliceAI.

Before You Begin

• Install OpenSpliceAI: Ensure you have installed OpenSpliceAI and its dependencies as described in the Installation page.

• Check Example Scripts: We provide an example script examples/predict/predict_cmd.sh

One-liner Start

OpenSpliceAI can predict splice sites from DNA sequences with a default 10,000 nt flanking region. You need:

1. A reference genome (FASTA) : chr22.fa

2. A reference annotation (GTF) : chr22.gff

3. A pre-trained OpenSpliceAI model or directory of models:

   • GitHub (models/spliceai-mane/10000nt/) or

   • FTP site (OSAI-MANE/10000nt/)

Run the following commands (adapt or replace filenames as needed):

openspliceai predict \
    -m 10000nt/ \
    -o results \
    -f 10000 \
    -i chr22.fa \
    -a chr22.gff \
    -t 0.9

This command will generate prediction results in the specified output directory (results/). The predictions will be based on the input FASTA file (chr22.fa) and the annotation file (chr22.gff). The results will include a GFF file with predicted splice sites and their scores.

Next Steps

• Explore ``predict`` Options: Dive into the predict documentation to learn more about the available options for predicting splice sites.

• Begin Variant Prediction: Check out the Quick Start Guide: variant guide to predict the impact of genomic variants on splice sites.

We hope this quick start guide helps you get up and running with OpenSpliceAI. Happy predicting!