pred-herg 5.0
Unlocking the Full Potential of Predictive Modeling for Precise hERG Activity Predictions
A Comprehensive Guide to Using the Pred-hERGApp's Advanced Machine Learning Model for Reliable Results
News
We are excited to introduce Pred-hERG 5.0, an innovative machine learning web application for predicting hERG blockage. This advanced tool offers enhanced accuracy, user-friendliness, and robustness to meet the needs of the global research community. Here are the key highlights of Pred-hERG 5.0:
Advanced Machine Learning
By harnessing the power of cutting-edge machine learning algorithms, we have taken significant strides in improving the accuracy of hERG blockage predictions, empowering researchers to make highly informed decisions with utmost confidence and precision.
Expansion of Database
Our integrated database has expanded significantly, now encompassing comprehensive information on over 14,364 compounds sourced from ChEMBL v30 database, providing an extensive and diverse array of data points for accurate predictions.
Meticulous Data Curation
To ensure the utmost accuracy and reliability, we have implemented a rigorous data curation process including removal of salts, mixtures, inorganic and organometallic compounds, normalization of chemotypes, and elimination of duplicates.
Diverse Prediction Models
Pred-hERG 5.0 introduces a robust framework consisting of three distinct models — classificatory, multi-classificatory, and regression — optimized to accommodate a wide range of prediction tasks.
Improved Prediction Accuracy
Pred-hERG 5.0 sets a new standard in performance, surpassing previous iterations and outshining other available tools with remarkable advancements in prediction accuracy rates.
User-Centric Design
With a commitment to addressing the needs of our user community, we have redesigned the interface for enhanced usability and intuitive navigation, making Pred-hERG accessible to researchers, students, and professionals.
PredHerg Predictor
INSTRUCTIONS
Draw a molecule
Use the molecular editor to draw the chemical structure of the compound you want to evaluate.
Hit the predict button and wait for results
Click 'PREDICT' to submit your molecule. Requests are processed via a fair queue. Please wait and do not close the page while your prediction is being processed.
Review consensus hERG blockage results
After prediction completes, you will see the consensus verdict (Blocker/Non-blocker), potency classification, and a breakdown of all three model predictions with their confidence scores.
Examine per-model breakdown and AD status
Review the individual predictions from the binary, multiclass, and regression models. Each model card shows its applicability domain status and similarity distance for reliability assessment.
Before you submit
Data handling, curation & privacy
A short, honest note on what PredHerg does and does not do with the chemistry you submit. Please read before running predictions on novel structures.
PredHerg does not perform full chemical standardization on submitted structures. Before predicting, please curate your inputs: remove counter-ions and salts, neutralize charges where appropriate, and select a canonical tautomer. The server applies only basic RDKit canonicalization on the SMILES it receives. Downstream predictions and applicability-domain checks are sensitive to the exact form you submit.
If RDKit cannot parse your structure (malformed SMILES, broken valences, exotic atoms outside the model's training distribution), the request fails with an error message. Always sanity-check inputs in a chemistry editor before submitting. For parseable but chemically unusual molecules, predictions are still returned. Use the Applicability Domain panel in the results to judge whether the prediction is trustworthy.
We do not persist your submitted SMILES, SDF, or MOL data. To accelerate repeated queries we cache results in Redis keyed only by a SHA-256 hash of the canonical SMILES, with a 7-day expiry. The original structure is never written to disk on our side. You can submit novel chemistry with confidence.
The free PredHerg web server predicts a single molecule per submission. SDF and MOL files are read in your browser by the chemistry editor, which loads one structure at a time onto the canvas, so any uploaded file is treated as a single-molecule request. Multi-molecule batch prediction will be available exclusively through our upcoming commercial offering, Insight AI Pro. For early access and pricing, please contact us at carolina@ufg.br.