Teppei Shimamura

This study introduces TISCA2 and IncepTIS to map and predict stress-responsive translation initiation sites, revealing tissue-specific patterns and enabling engineered mRNA 5' leaders.

mmVelo estimates cell state-dependent dynamics across multiple modalities using deep generative modeling, validated on mouse and human data.

MUSE integrates multi-omics data across species to align cellular states while preserving species-specific features and regulatory module conservation.

UNISON is a scalable framework for single-cell transcriptomics that uses skip-zeros variational inference to efficiently analyze large, sparse datasets while preserving interpretability.

scTREND is a hazard model that estimates cell-level risk across time and conditions without cell-type annotations, showing improved prediction in simulations and applications to cancer and COVID-19.

scSurv estimates individual cellular contributions to cancer patient survival using a deep generative model combined with Cox proportional hazards analysis.

STODE reconstructs continuous spatiotemporal dynamics from discrete spatial transcriptomics data, outperforming existing methods in predicting gene expression and spatial coordinates.

LineageVAE reconstructs historical cell states and transcriptomes from scRNA-seq data to infer unobserved progenitor cell states.

This paper introduces an unsupervised deep learning method for deblurring, super-resolution, and segmentation of two-photon microscopy images, achieving better performance than conventional methods.

The paper introduces LincSpectr, a deep generative model that integrates transcriptomic and electrophysiological data to uncover molecular mechanisms underlying time-frequency neural activities.

DeepCOLOR identifies cell-cell interactions by reconstructing spatial colocalization maps from single-cell and spatial transcriptomes in various tissues.