Jan Lause

This paper introduces a compound model for single-cell RNA-seq data without UMIs, using Pearson residuals to improve gene selection and embeddings in Smart-seq2 datasets.

Eyewire II provides a large-scale EM dataset of the mouse retina, enabling cell type annotation and connectivity mapping through human-in-the-loop deep learning and functional alignment.

The paper challenges claims that t-SNE and UMAP embeddings of single-cell data are misleading, showing they can still provide biologically relevant information despite not preserving high-dimensional distances.

The paper identifies two distinct synaptic sites on horizontal cells in the mouse retina that mediate local feedback to cones and global feedforward signaling to bipolar cells.

This paper compares normalization methods for single-cell RNA-seq data, showing that analytic Pearson residuals outperform other approaches in identifying variable genes and dimensionality reduction.