Pathogenicity Predictor User Guide

Pathogenicity Predictor is a mutation pathogenicity prediction tool based on bioinformatics and AI models. It can predict whether single-base mutations (including insertions and deletions) in the human genome are likely to cause serious diseases, providing efficient and reliable pathogenicity analysis support for research and clinical practice. It is suitable for researchers and clinicians studying the association between gene mutations and diseases, addressing the low efficiency and insufficient accuracy of traditional methods.

Open a browser and navigate to: https://www.icyagen.com/tools/pathogenicity to access the Pathogenicity Predictor homepage.

The tool supports 4 input modes, including: Paste, Mutation, Protein, and Transcript.

• Fill in the gene name: Enter "SOD1" in the "Gene Name" input box. The system will automatically suggest recommended genes; select the target gene.
• Select the transcript: The system defaults to the most commonly used transcript (e.g., SOD1:NM_000454.5). Users can select other transcripts from the dropdown menu if needed.

• Fill in the mutation site: In the "Amino acid Site" field, enter the protein mutation site. For example, in the mutation c.335G>T, the 335th nucleotide being G typically codes for the 112th amino acid, which the system automatically resolves to Cys (C): TGC.
• Select the post-mutation base according to research needs. The mutated nucleotide is marked in red, while the non-mutated ones are black.

• Visualize mutation site: Click the [Unfold Diagram] button below the input box to see the mutation's location in the gene structure.
• Add or submit mutations: To analyze multiple sites, click the [Add a mutant] button and repeat the steps.
• RNA Splicing Prediction: Manually enable via checkbox. Once activated, the result page will additionally display the impact of variants on splicing sites, including prediction scores and supporting evidence.
• After confirming that the information is correct, click the [Submit] button at the bottom of the page to start the AI model prediction. After submission, a progress bar will be displayed on the page. The prediction time depends on the number of mutations and the server load.

• Clinical Significance Judgment: Directly displays the pathogenicity classification of the mutation (e.g., "Likely Pathogenic," "Uncertain Significance," "Likely Benign") based on a comprehensive analysis of AI models and databases (like ClinVar, OMIM).
• Pathogenicity Quantitative Index: Displays a score between 0 and 1, representing the probability of pathogenicity (e.g., a score of 0.6321 = 63.21% pathogenic probability). The higher the score, the higher the pathogenic risk.

In the [Sequence Details] page, you can view a comparison of the base/amino acid sequences before and after the mutation, with the mutation site highlighted, to help assess the mutation's impact on protein structure.

If you enable Concurrent RNA splicing prediction, the results page will also show the mutation's effect on splice sites, complete with prediction scores and supporting evidence.

The page supports saving results as screenshots and allows for downloading to a local file or exporting to an email address.

• Gene Naming Standardization: Enter standard gene symbols (e.g., "TP53" instead of "p53"). Correct names can be verified via NCBI Gene.
• Mutation site Format: Follow HGVS nomenclature (e.g., "c.123A>T" for coding sequence mutations, "g.456C>G" for genomic mutations). Incorrect formatting may cause parsing errors.
• Transcript Selection: The tool defaults to MANE transcripts. Ensure the correct transcript ID is selected based on the mutation source, as incorrect selection may lead to positional errors.
• Multi-Mutation Analysis: A maximum of 5 mutation sites can be submitted at once. For batch analysis, submit in multiple rounds.

• Codon Chart: A codon table is available on the left to help understand the impact of mutations.
• Mutation Formatting Rules: The question mark icon next to [Standard Mutation Nomenclature] provides formatting guidelines to help ensure correct syntax.

• Research: Predicting the pathogenic mechanisms of disease-related gene mutations and analyzing their functions.
• Clinical: A support system for assessing the pathogenicity of genetic mutations based on ACMG guidelines.
• Education: A dynamic, visual teaching solution for demonstrating the impact of mutations on protein structure and function.
• Industry: A precision filtering strategy for high-frequency, harmless mutations in drug target screening.