Large-scale neural datasets are becoming increasingly available, from gene expression to morphology and spiking activity. Learning aligned representations of these data may help us study how these modalities relate and enable joint analyses.
Azhar Akhmetova
Azhar works on multimodal integration of datasets describing neuronal populations from the same brain region. In particular, she focuses on transcriptomic, electrophysiological, and morphological data. These modalities provide complementary views of the underlying biology and may therefore have shared latent structure. Her goal is to bring these views together in a shared representation space. To achieve this, she builds encoder models that jointly embed and align modalities in an unsupervised way using optimal transport, which compares the geometric structure of each modality. She is a PhD student and a member of the IMPRS-IS graduate school.
Large-scale neural datasets are becoming increasingly available, from gene expression to morphology and spiking activity. Learning aligned representations of these data may help us study how these modalities relate and enable joint analyses.
Present Positions And Title
PhD Student
Research Group
Email Address
Academic Education
| Year | Degree | Institution | Field of Study |
|---|---|---|---|
| 2022 - 2026 | MSc | University of Göttingen | Physics |
| 2018-2022 | BSc | Nazarbayev University | Physics |