{% extends "base.html" %} {% block title %}Tutorial - {{ settings.app_name }}{% endblock %} {% block content %}

Tutorial

A step-by-step guide to predicting RNA splicing outcomes

This tutorial takes about 5 minutes to complete
1

Prepare the Sequence

A 70-nucleotide exon sequence is required. This represents the cassette exon to be analyzed.

Checklist

  • Exactly 70 nucleotides long
  • Only A, C, G, T (no U, N, or other characters)
  • No spaces, headers, or line breaks

Don't have a sequence?

No problem! Click the "Try Example" button on the home page to load a sample sequence.

2

Enter the Sequence

Go to the home page and paste the sequence into the text area.

Example Sequence

GGTAGTACGCCAATTCGCCGGTGCCGCGAGCCAGAGGCTACCAAAACTTGACAAGCCTACATATACTACT

This sequence has high inclusion (PSI ~0.98)

While typing, the interface will show:

  • • A character counter showing X/70 nucleotides
  • • Validation messages if something is wrong
  • • The "Predict PSI" button will enable when the input is valid
3

Submit and Wait

Click the "Predict PSI" button to submit the sequence for analysis.

What happens behind the scenes

  1. 1 10nt flanking sequences are added to create a 90nt window
  2. 2 ViennaRNA predicts the secondary structure
  3. 3 The neural network computes PSI and position contributions
  4. 4 Results are saved and the user is redirected to the result page

Typical prediction time: 2-5 seconds

4

Understand the Results

The result page shows several pieces of information:

PSI Value

The main prediction, shown as a large number (0.00 to 1.00).

0.8-1.0
High Inclusion
0.3-0.8
Variable
0.0-0.3
High Skipping

RNA Secondary Structure

Shown in dot-bracket notation:

. = unpaired, ( = base paired (5' side), ) = base paired (3' side)

Minimum Free Energy (MFE)

The thermodynamic stability of the predicted structure. More negative = more stable.

Force Plot

Shows how each position contributes to the PSI:

Promotes inclusion
Promotes skipping
5

Export Results

Results can be downloaded in multiple formats:

CSV

For spreadsheets (Excel, Google Sheets)

JSON

For programmatic analysis

Link

Share or bookmark results

Ready to Try?

Head to the home page and make a first prediction!

Start Predicting

Using the API

For programmatic access, the REST API can be used directly.

Quick Example (Python)

import requests

# Submit a prediction
response = requests.post(
    "{{ request.url.scheme }}://{{ request.url.netloc }}/api/predict",
    json={"sequence": "GGTAGTACGCCAATTCGCCGGTGCCGCGAGCCAGAGGCTACCAAAACTTGACAAGCCTACATATACTACT"}
)

result = response.json()
print(f"PSI: {result['psi']}")
print(f"Structure: {result['structure']}")

Quick Example (curl)

curl -X POST "{{ request.url.scheme }}://{{ request.url.netloc }}/api/predict" \
  -H "Content-Type: application/json" \
  -d '{"sequence": "GGTAGTACGCCAATTCGCCGGTGCCGCGAGCCAGAGGCTACCAAAACTTGACAAGCCTACATATACTACT"}'

See the full API documentation for more endpoints and options.

Help & FAQ API Documentation
{% endblock %}