Bansal, I. | Search bioRxiv

bioRxiv2026

Machine learning assisted multi-level differentiation of blood cells using multi-frequency electro-mechanical impedance cytometry

We report a multi-frequency microfluidic impedance cytometer with multimodal measurement of electrical and mechanical cell properties. Using the large feature set from the multimodality, we implemented machine learning assisted differentiation of cells. We have demonstrated a constriction-equipped channel to study the mechanical cell properties, and integrated multiple microelectrodes that capture electric field information pertaining to various stages of cell transit. Our system works over a wide frequency range from 800 kHz to 8 MHz. We developed a machine learning workflow to classify cells using our multi-modal data. Using this platform to show blood cell differentiation, our system shows minimum F1 score to be 0.92, minimizing both false negatives and false positives, which is important for medical diagnosis. Furthermore, we also identified the presence of two populations in the mixed sample of neutrophils and PBMCs. Our workflow allows classification of cells in a new sample along with identification of top 5 distinguishing parameters out of 70+ features extracted from the raw impedance signal.

bioengineering↗

bioRxiv2025

The discovery and patterns of the underlying long-standing mild-effect mutator alleles in S. cerevisiae populations

Most mutations are neutral or deleterious, and mutator alleles that substantially increase the mutation rate of an organism are believed to be short-lived in natural populations. Theory suggests that mutator alleles of modest effect have the potential to segregate nearly neutrally, but this prediction has been difficult to test experimentally. Here, we report strong genomic signatures consistent with the transmission and long-term maintenance of at least one natural mutator allele in Saccharomyces cerevisiae isolates. Specifically, such signatures are consistent with segregation of mutator alleles that disproportionately increase A>C mutations in the African Beer strains of S. cerevisiae. Remarkably, the mutation spectra of African Beer strains deviate from that of other S. cerevisiae natural isolates more than some Saccharomyces species differ from each other. Furthermore, computational analysis suggests mutator allele introgression from the African Beer population into a subset of French Dairy yeast, motivating us to experimentally measure the de novo mutation rates and spectra of these strains. We observed a consistent but weak enrichment of A>C and A>G de novo mutations in the African Beer population, with the A>C enrichment also evident in the French Dairy subset. Unexpectedly, other de novo mutation types varied more substantially among strains, including one outlier African Beer strain (AFL) with an excess of C>A mutations. Overall, our measurements reveal that de novo mutation spectra show greater variability among populations compared to the mutation spectra of rare polymorphisms, which in turn show greater variability than the spectra of all polymorphisms. Together, while additional mutator alleles may segregate in natural populations, mild enrichment of A>C and A>G types in African Beer strains reflects mutator alleles with weak effects that can persist and accumulate mutations through evolution.

evolutionary biology↗