Anders Krogh

Extrachromosomal circular DNA formation follows a 10.5-bp periodicity reflecting the minor groove orientation around nucleosomes

The paper analyzes eccDNA length distributions and breakpoint organization across multiple datasets to show that eccDNA formation is influenced by nucleosome positioning and DNA helix orientation.

A deep generative decoder predicts microRNA expression from bulk and single-cell mRNA profiles

miDGD predicts microRNA expression from bulk and single-cell mRNA profiles using a deep generative decoder, achieving high accuracy across diverse datasets and cell types.

DeepTMHMM2 enables accurate prediction of transmembrane protein topology and subcellular location

DeepTMHMM2 predicts transmembrane protein topology and subcellular location, including re-entrant regions and interfacial helices, and identifies localization across 17 membranes.

A generative model for dimensionality reduction with millions of features and few samples

This paper introduces a deep generative decoder model for dimensionality reduction with millions of features and few samples, showing it outperforms PCA and VAE in tumor classification.

Enhancing Generative Decoders with Stochastic Training on Biomedical Data with Missingness

This paper introduces stochastic training strategies to improve deep generative decoders for biomedical datasets with missing values, showing improved robustness and convergence.

multiDGD: A versatile deep generative model for multi-omics data

multiDGD is a deep generative model for multi-omics data that learns shared representations and enables data integration and association detection.

N-of-one differential gene expression without control samples using a deep generative model

This paper introduces a deep generative model to perform differential gene expression analysis on single cancer samples without control samples, outperforming DESeq2 in detecting differentially expressed genes.

Prediction of DNA from context using neural networks

The paper trains a neural network to predict DNA bases from context, achieving 54% accuracy on the human genome, outperforming Markov models and revealing periodic signals linked to nucleosome positioning.

DeepTMHMM predicts alpha and beta transmembrane proteins using deep neural networks

Combining an ESM-1b encoder with a conditional random field decoder, DeepTMHMM classifies alpha-helical and beta-barrel transmembrane proteins and their topology, outperforming most compared tools, including on signal-peptide cleavage sites.

Context dependency of nucleotide probabilities and variants in human DNA

The paper develops context-dependent nucleotide models for the human genome, showing they can predict bases with over 50% accuracy and fit somatic variants better than simpler models.

Taxonomic classification method for metagenomics based on core protein families with Core-Kaiju

The paper introduces Core-Kaiju, a taxonomic classification method for metagenomics that improves accuracy in predicting taxa number and abundance compared to existing methods.

CAncer bioMarker Prediction Pipeline (CAMPP) - A standardised and user-friendly framework for the analysis of quantitative biological data.

CAMPP is an open-source R pipeline that simplifies high-throughput data analysis for non-experts, providing tools for normalization, clustering, differential expression, and biomarker prediction.

Sensitive detection of circular DNA at single-nucleotide resolution using guided realignment of partially aligned reads

Circle-Map improves detection of circular DNA from short-read sequencing data by realigning partially aligned reads using discordantly mapped reads.

Kaiju: Fast and sensitive taxonomic classification for metagenomics

Kaiju is a metagenomic classifier that improves sensitivity over k-mer-based methods by using protein-level exact matches and optional amino acid substitutions.