Emanuel Michele Soda

Computational Scientist · Spatial Transcriptomics · Machine Learning

Milan, Italy  ·  emanuelsoda@gmail.com  ·  github.com/EmanuelSoda  ·  linkedin


Summary

Computational scientist with 5+ years of experience developing machine learning, Bayesian statistical models, and reproducible data pipelines for multimodal biomedical data, including spatial transcriptomics, single-cell genomics, digital pathology, live cell imaging, and clinical outcome modelling.


Skills

Programming & Data

PythonRSQLBash/LinuxGit/GitHubDocker

Machine & Deep Learning

PyTorchScikit-learnstatsmodelsPyroscvi-toolsTidymodels

Bioinformatics

ScanpySeuratSquidpySpatialDataNextflowHPC

Image & Spatial Analysis

SegmentationRegistrationOpenCVscikit-imageLazySlideImageJ

Experience

Visiting PhD Student

Wellcome Sanger Institute, Hinxton, UK — Bayraktar group

Designed multimodal deep learning frameworks integrating GeoMx RNA profiles and H&E histology features to characterise crypt-level heterogeneity across IBD and healthy control patients.

Doctoral Researcher

Human Technopole & Politecnico di Milano, Milan, Italy

PhD in Data Analytics and Decision Sciences.

  • Developed reproducible computational pipelines integrating spatial transcriptomics, histopathology, cell type composition, and clinical outcomes.
  • Characterised spatially resolved tissue microenvironments across 180 IBD and healthy control samples using probabilistic deconvolution and NMF-based niche discovery.
  • Built Bayesian classification and regression models linking spatial composition to clinical outcomes for interpretable biomarker discovery from Visium data.
  • Developed computational pathology workflows integrating instance segmentation and self-supervised representation learning to quantify crypt-level morphological heterogeneity across IBD whole slide images.

Bioinformatician & ML Scientist

Istituto Europeo di Oncologia (IEO), Milan, Italy

  • Developed and validated an automated ML pipeline for cell cycle phase classification from live cell imaging data — co-authored eLife publication.
  • Analysed bulk and single-cell AML data linking metabolic to epigenetic alterations.
  • Reduced manual processing time by 80% through automated computational analysis.

Education

PhD in Data Analytics and Decision Sciences

Human Technopole & Politecnico di Milano, Milan, Italy

MSc in Bioinformatics for Computational Genomics

Politecnico di Milano & University of Milan, Milan, Italy

Developed ScreenR, an R/Bioconductor package for automated hit identification in shRNA loss-of-function screens, integrating edgeR-based statistical testing within a Tidyverse-native pipeline.

BSc in Computer Science

University of Milan Bicocca, Milan, Italy


Publications

Sara Ometto, Soumick Chatterjee, …, Emanuel M. Soda, …, Craig A. Glastonbury. Hundreds of cardiac MRI traits derived using 3D diffusion autoencoders share a common genetic architecture. Nature Communications, 2026.

Kourosh Hayatigolkhatmi*, Chiara Soriani*, Emanuel M. Soda* et al. Automated workflow for the cell cycle analysis of (non-)adherent cells using a machine learning approach. eLife, 2024.


Awards & Grants

1st Place — Time-Series Exploration with Deep Learning Hackathon

Milan, September 2024

Best Project Award — Bioinformatics Hackathon (Protein Edition)

Copenhagen, July 2021


Teaching & Mentoring

Master's thesis co-supervision (×2)

Bioinformatics for Computational Genomics, Milan

Introduction to Computational Biology — Seminar

Molecular pharmacology & immunology, University of Milan