Recent News
Graduation 2024
Congratulations to Haidong and Sneha who graduated on Saturday May 12. Haidong completed his PhD and Sneha completed her BS. We have been so grateful to have them in the lab for the past few years. Haidong was the first member of the lab and notably developed CytoSet, which is a set-based encoding strategy for featurizing multi-sample single-cell datasets. We have been able to readily use CytoSet across a range of single-cell datasets where we to predict an outcome from the entire immune landscape. Haidong’s commitment to writing good software and to setting up important pipelines in the lab has been invaluable. Thank you Haidong for all of your hard work and creativity! Sneha finished her BS in computer science and minored in neuroscience. She also earned highest honors in computer science, based on her work in studying age-related changes in the microglia transcriptome. Sneha’s bold curiosity has been quite impressive, and it was an absolute joy to work with her.
Celebrating the End of the Spring 2024 Semester
Thanks to all the amazing students in the lab / collaborators for a productive and fun semester and year! Here are some highlights of new papers and proposals.
DELVE. Jolene’s method DELVE for trajectory-preserving feature selection was published in nature communications. We are already using this in many projects with dynamic or temporal single-cell data to identify modules of features with related dynamic patterns.
Differential Abundance Benchmarking. Haidong’s differential abundance benchmarking study was published in Genome Biology. This paper systematically explores how to choose the appropriate differential abundance method for a given single-cell dataset to most readily identify condition-specific prototypical cells.
Funding. We got an R21 funded in collaboration with Todd Cohen to develop new methods for identifying meaningful spatial configurations of cell-types and pathologies in the brain that correlate with the aging and neurodegeneration process, as assayed through imaging mass cytometry. Very excited to continue to develop the neuroimmune focus in the lab.
New Student. We are very excited to welcome a new BCB student to the lab, Luvna Dhawka. Luvna has a very interesting background in neuroscience and computational biology and will use these expertise to study sex differences in the microglia transcriptome in aging and neurodegeneration. Welcome Luvna!
Very proud of all of our students for all of the hard work this semester!!
2023 was a great year!
Thanks to our great students, collaborators, and colleagues for another great year. This was a great year of starting to work more in neuroimmunology, graduating our first PhD student, and starting new collaborations at UNC. Here is a photo from a recent lab dinner as osteria georgi (highly recommend. I am not really a pasta person, but this was impressive).
NIH Grant! R21 Funded to Integrate Data Across Multiple Ex-Vivo Stimulations. (August 2023)
Big thanks to NIAID for funding our first NIH grant to develop new methods to identify and functionally characterize immune cells subjected to ex-vivo stimulation. As part of the project, we will be generating many publicly available datasets of stimulation specific immunological features that can be used for downstream predictions. Read more about the project here!
Welcome DELVE (May 2023)
We are excited to introduce our new biological trajectory preserving feature selection approach called DELVE . As we know, it is often essential to correctly establish an ordering of cells along some kind of developmental or differentiation trajectory. However, single-cell measurements can become complicated when there are a large number of measured features, and these features are noisy. This is particularly the case in high-dimensional single-cell assays, such as scRNA-seq, where there are ~20K genes measured per-cell. DELVE provides a principled approach for feature selection, where we specifically aim to identify dynamic modules of features that are prototypical of some point along the developmental trajectory. We applied DELVE across diverse applications, including, CD8 T-cell differentiation and the cell-cycle.
Congratulations Dr. Jolene Ranek! (March 2023)
Congratulations to Dr. Jolene Ranek who successfully defended her PhD thesis entitled ‘Resolving Biological Trajectories in single-cell data using feature selection and multi-modal integration’. Jolene has made significant contributions in bioinformatics for RNA velocity, and in trajectory-preserving feature selection techniques. We will certainly miss her a whole lot when she leaves for her postdoc in a few months (we’ll let her share this news ;) ) . Congratulations Jolene!!
Photo from the doctoral hooding ceremony in May, 2023.
Two Pilot Awards from UNC Generously Funded! (December 2022)
We are grateful to have received two pilot awards for two different projects related to T-cell Differentiation and Imaging CyTOF for Alzheimer’s Disease from Computational Medicine and Neuroscience, respectively. Thanks for the support and for helping us to work with our awesome collaborators Todd and Justin!!
Very happy to Collaborate with Justin Milner on the computational medicine pilot award “A Dual Screening Approach for Engineering Durable and Effective T-Lymphocyte Responses”
Also very happy to collaborate with Todd Cohen on a NeuroSpark pilot award “Spatial signatures of brain health and vulnerability in aging and Alzheimer’s disease”
RNA Velocity + Expression Integration Benchmarking Study Published in Genome Biology (September 2022)
Congratulations to Jolene and Jeremy for a new paper in Genome Biology on a comprehensive evaluation of biological multimodal integration approaches for for combining RNA velocity and expression information from single cells. We evaluated different ways of combining expression and velocity information for clinical prediction tasks and trajectory-preserving metrics and somewhat unsurprisingly found that often simple integration strategies (such as concatenation!) can be quite effective for integrating complementary expression and velocity information.
Cytocoarsening Paper Accepted to PSB 23 (September 2022)
Congratulations to Chi-Jane and Emma whose paper A Graph Coarsening Algorithm for Compressing Representations of Single-Cell Data with Clinical or Experimental Attributes was accepted to PSB23! Given that many methods for linking cellular heterogeneity to clinical phenotype are graph-based, we developed a graph coarsening approach based on both graph structure and clinical attributes to make graphs smaller without losing performance in downstream tasks. Science aside, we are looking forward to hanging out our colleagues on Hawaii. :)
Papers at ACM BCB 2022 (August 2022)
Congratulations to Haidong, Jolene, Siyuan, Vishal, and Junier for papers recently published at ACM-BCB
CytoEMD : Paper
Distribution Preserving Sketching : Paper
Transparent Single-Cell Sample Classification with Kernel Mean Embeddings : Paper
We were thrilled and grateful to find out that our paper on Single-cell set classification with Kernel mean embeddings won the best paper award! Big congratulations to Siyuan, Vishal, Haidong, Jolene, and Junier!
Older News (From the beginning of the lab (January 2021) until May 2022)
Hello World (January 2021)
Natalie joins UNC CS + CompMed as an assistant professor. Haidong joins, followed by Jolene.
CytoSet Accepted to ACM BCB 2021 (June 2021)
Congratulations to Haidong, whose CytoSet paper was accepted to ACM BCB 2021. Code for Cytoset is available here.
Welcome New Lab Members! (August 2021)
During the fall semester we were happy to welcome new graduate students Chi-Jane Chen, and Alec Plotkin (BCB Rotation student) and undergraduate student Nora Xia. Chi-Jane, Alec, and Nora are all working on some aspects of graphs for single-cell data. Alec is helping to teach us more immunology and getting us excited about T-cells.
Review Article in Nature Immunology (Sept 2021)
We were honored to contribute to a review article in nature immunology about best practices for the analysis of Cytometry data. Thanks Florian for leading the effort! [paper]
New Paper on ArXiv on Embedding Multiple Sets of Cells in an Interpretable Manner (January 2022)
Towards this challenging problem of understanding which type of cells differ between patients across phenotypes, our recent work in collaboration with Siyuan Shan and Junier Oliva computes random Fourier features for each cell and represents each sample’s set of cells as the mean in the random Fourier feature space. This approach not only allows us to classify patient samples quite accurately according to phenotype, but allows us to understand the cells driving differences between classes. Congratulations to Siyuan for this hard work. [ArXiv Pre-Print]
Alec Joins the Lab (May, 2022) !
After his three rotations in BCB, we are thrilled that Alec joined us officially to continue and build the collaboration with his co-mentor Justin Milner. :)