# CellMateA Deep Learning-Assisted Single-Cell Data Processing Platform

**Authors:** Felix Friedrich, Cátia Marques, Ingela Lanekoff

PMC · DOI: 10.1021/acs.analchem.5c07205 · Analytical Chemistry · 2026-02-13

## TL;DR

CellMate is a new platform for processing single-cell metabolomics data using deep learning, making it easier to analyze complex datasets from individual cells.

## Contribution

CellMate introduces a deep learning-based image classification algorithm for untargeted metabolomics workflows in single-cell data.

## Key findings

- CellMate provides a user-friendly interface for metabolite identification and peak alignment in single-cell data.
- The deep learning algorithm effectively separates endogenous metabolites from background noise.
- CellMate is compatible with direct infusion techniques and supports both targeted and untargeted workflows.

## Abstract

Mass spectrometry-based single-cell metabolomics (SCM)
reveals
the inherent heterogeneity of individual cells among seemingly identical
cell types. Fast-scanning and high-resolving mass analyzers provide
the sensitivity and specificity required to probe minuscule amounts
of biological material. However, acquiring data from hundreds of individual
cells to achieve statistical power results in complex data sets. This
challenge is compounded by the limited availability of specialized
data analysis tools for single-cell metabolomics, as many techniques
depend on the use of specialized sampling and ionization probes. This
results in incompatibility with conventional metabolomics data processing
tools. Here, we present CellMate, a MATLAB-based data processing platform
designed for single-cell metabolomics using direct infusion techniques.
CellMate comprises identification and peak alignment of detected metabolites
in an intuitive graphical user interface. CellMate supports customizable
quantitative, targeted, and nontargeted metabolomic workflows. The
untargeted workflow is enabled by a novel deep learning-based image
classification algorithm that effectively distinguishes endogenous
metabolites from background species. The source code, along with a
compiled installer, is available at github.com /LanekoffLab/CellMate. We believe that CellMate
represents a significant advancement in the single-cell metabolomics
toolbox, enabling comprehensive data extraction of precious metabolite
information from single cells.

## Full-text entities

- **Diseases:** malaria (MESH:D008288), prostate cancer (MESH:D011471), DL (MESH:D007859), SCM (MESH:D012640)
- **Chemicals:** S (MESH:D013455), Ag107 (-), amino acid (MESH:D000596), arginine (MESH:D001120), PC (MESH:C053518), TCA (MESH:D014238), lipids (MESH:D008055), H (MESH:D006859), PA (MESH:D011478), metal (MESH:D008670), salt (MESH:D012492), formic acid (MESH:C030544), N (MESH:D009584), choline (MESH:D002794), Valine (MESH:D014633), H2O (MESH:D014867), ACN (MESH:C084683), silver (MESH:D012834), oleamide (MESH:C029407), 13C (MESH:C000615229)
- **Species:** Homo sapiens (human, species) [taxon 9606]
- **Cell lines:** S2 — Drosophila melanogaster (Fruit fly), Spontaneously immortalized cell line (CVCL_Z232)

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC12937054/full.md

## Figures

5 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12937054/full.md

## References

44 references — full list in the complete paper: https://tomesphere.com/paper/PMC12937054/full.md

---
Source: https://tomesphere.com/paper/PMC12937054