Machine Learning

Computational Drug Discovery Part 3 (Subpart 2/3): AlphaFold's Evoformer Block Disassembled, A Matrix-Level Deep Dive into AlphaFold2's Core

A detailed mathematical breakdown of AlphaFold2’s Evoformer block, explaining each operation with concrete matrix algebra and dimensions - Subpart 2/3

Diagram showing efficient transformer architectures

Taming the Transformer: How Perceiver IO and PaCa-ViT Conquer Quadratic Complexity

A deep dive into two novel architectures, Perceiver IO and PaCa-ViT, that break the O(N^2) barrier in Transformers, enabling them to process massive inputs efficiently.

Computational Drug Discovery Part 4: Graph Neural Networks for Molecular Property Prediction

A technical deep-dive into Graph Neural Networks (GNNs) for predicting molecular properties. Learn how to construct molecular graphs, implement message passing architectures, and apply attention mechanisms to drug discovery tasks.

Computational Drug Discovery Part 3 (Subpart 1/3): AlphaFold Overview

How DeepMind’s AlphaFold2 solved the 50-year grand challenge in biology – the protein folding problem – using transformers, evolutionary information, and geometric reasoning and what it means for drug discovery - Subpart 1/3

Computational Drug Discovery Part 2: Molecular Representations for Machine Learning

Molecules are converted into data for ML using SMILES, fingerprints, molecular graphs, and 3D geometry. The trade-off is between compactness (SMILES/fingerprints) and richness (graphs/3D geometry). The representation choice fundamentally determines what models can learn about binding and performance in drug discovery.

Computational Drug Discovery Part 1: Molecules, Proteins, and the Drug Discovery Challenge

An introduction to the fundamental biology behind drug discovery. We explore atoms, bonds, functional groups, and proteins, then examine the computational drug discovery pipeline from target identification to clinical trials. Learn why finding effective drugs is extraordinarily difficult and how AI is transforming this challenge.

Reliability of atrial fibrillation detection

Reliability of Self-Applied Smartphone Mechanocardiography for Atrial Fibrillation Detection

This study investigates the reliability of self-applied smartphone mechanocardiography (sMCG) for the detection of atrial fibrillation (AFib). The results show that sMCG can accurately differentiate AFib from sinus rhythm in both physician- and self-applied recording scenarios.

Comprehensive Analysis of Cardiogenic Vibrations for Automated Detection of Atrial Fibrillation Using Smartphone Mechanocardiograms

This paper presents a comprehensive time-frequency pattern analysis approach for automated detection of AFib from smartphone-derived seismocardiography (SCG) and gyrocardiography (GCG) signals. The experimental results showed high accuracy, sensitivity, and specificity for both cross-validation and cross-database tests.

An Activity Recognition Framework Deploying the Random Forest Classifier and A Single Optical Heart Rate Monitoring and Triaxial Accelerometer Wrist-Band

This paper investigates a range of daily life activities and uses a random forest classifier to detect them based on wrist motions and optical heart rate. The highest accuracy was achieved with a forest of 64 trees and 13-s signal segments.