Search for: All records

Dictionary learning (DL), implemented via matrix factorization (MF), is commonly used in computational biology to tackle ubiquitous clustering problems. The method is favored due to its conceptual simplicity and relatively low computational complexity. However, DL algorithms produce results that lack interpretability in terms of real biological data. Additionally, they are not optimized for graph-structured data and hence often fail to handle them in a scalable manner. In order to address these limitations, we propose a novel DL algorithm calledonline convex network dictionary learning(online cvxNDL). Unlike classical DL algorithms, online cvxNDL is implemented via MF and designed to handle extremely large datasets by virtue of its online nature. Importantly, it enables the interpretation of dictionary elements, which serve as cluster representatives, through convex combinations of real measurements. Moreover, the algorithm can be applied to data with a network structure by incorporating specialized subnetwork sampling techniques. To demonstrate the utility of our approach, we apply cvxNDL on 3D-genome RNAPII ChIA-Drop data with the goal of identifying important long-range interaction patterns (long-range dictionary elements). ChIA-Drop probes higher-order interactions, and produces data in the form of hypergraphs whose nodes represent genomic fragments. The hyperedges represent observed physical contacts. Our hypergraph model analysis has the objective of creating an interpretable dictionary of long-range interaction patterns that accurately represent global chromatin physical contact maps. Through the use of dictionary information, one can also associate the contact maps with RNA transcripts and infer cellular functions. To accomplish the task at hand, we focus on RNAPII-enriched ChIA-Drop data fromDrosophila MelanogasterS2 cell lines. Our results offer two key insights. First, we demonstrate that online cvxNDL retains the accuracy of classical DL (MF) methods while simultaneously ensuring unique interpretability and scalability. Second, we identify distinct collections of proximal and distal interaction patterns involving chromatin elements shared by related processes across different chromosomes, as well as patterns unique to specific chromosomes. To associate the dictionary elements with biological properties of the corresponding chromatin regions, we employ Gene Ontology (GO) enrichment analysis and perform multiple RNA coexpression studies. 

Controlled trapping of cells and microorganisms using substrate acoustic waves (SAWs; conventionally termed surface acoustic waves) has proven useful in numerous biological and biomedical applications owing to the label- and contact-free nature of acoustic confinement. However, excessive heating due to vibration damping and other system losses potentially compromises the biocompatibility of the SAW technique. Herein, we investigate the thermal biocompatibility of polydimethylsiloxane (PDMS)-based SAW and glass-based SAW [that supports a bulk acoustic wave (BAW) in the fluid domain] devices operating at different frequencies and applied voltages. First, we use infrared thermography to produce heat maps of regions of interest (ROI) within the aperture of the SAW transducers for PDMS- and glass-based devices. Motile Chlamydomonas reinhardtii algae cells are then used to test the trapping performance and biocompatibility of these devices. At low input power, the PDMS-based SAW system cannot generate a large enough acoustic trapping force to hold swimming C. reinhardtii cells. At high input power, the temperature of this device rises rapidly, damaging (and possibly killing) the cells. The glass-based SAW/BAW hybrid system, on the other hand, can not only trap swimming C. reinhardtii at low input power, but also exhibits better thermal biocompatibility than the PDMS-based SAW system at high input power. Thus, a glass-based SAW/BAW device creates strong acoustic trapping forces in a biocompatible environment, providing a new solution to safely trap active microswimmers for research involving motile cells and microorganisms. 

Acoustic microfluidics has emerged as a versatile solution for particle manipulation in medicine and biology. However, current technologies are largely confined to specialized research laboratories. The translation of acoustofluidics from research to clinical and industrial settings requires improved consistency and repeatability across different platforms. Performance comparisons will require straightforward experimental assessment tools that are not yet available. We introduce a method for characterizing acoustofluidic devices in real-time by exploiting the capacity of swimming microorganisms to respond to changes in their environment. The unicellular alga, Chlamydomonas reinhardtii , is used as an active probe to visualize the evolving acoustic pressure field within microfluidic channels and chambers. In contrast to more familiar mammalian cells, C. reinhardtii are simple to prepare and maintain, and exhibit a relatively uniform size distribution that more closely resembles calibration particles; however, unlike passive particles, these motile cells naturally fill complex chamber geometries and redistribute when the acoustic field changes or is turned off. In this way, C. reinhardtii cells offer greater flexibility than conventional polymer or glass calibration beads for in situ determination of device operating characteristics. To illustrate the technique, the varying spatial density and distribution of swimming cells are correlated to the acoustic potential to automatically locate device resonances within a specified frequency range. Peaks in the correlation coefficient of successive images not only identify the resonant frequencies for various geometries, but the peak shape can be related to the relative strength of the resonances. Qualitative mapping of the acoustic field strength with increasing voltage amplitude is also shown. Thus, we demonstrate that dynamically responsive C. reinhardtii enable real-time measurement and continuous monitoring of acoustofluidic device performance. 

Functional cilia and flagella are crucial to the propulsion of physiological fluids, motile cells, and microorganisms. Motility assessment of individual cells allows discrimination of normal from dysfunctional behavior, but cell-scale analysis of individual trajectories to represent a population is laborious and impractical for clinical, industrial, and even research applications. We introduce an assay that quantifies swimming capability as a function of the variation in polar moment of inertia of cells released from an acoustic trap. Acoustic confinement eliminates the need to trace discrete trajectories and enables automated analysis of hundreds of cells in minutes. The approach closely approximates the average speed estimated from the mean squared displacement of individual cells for wild-type Chlamydomonas reinhardtii and two mutants ( ida3 and oda5 ) that display aberrant swimming behaviors. Large-population acoustic trap-and-release rapidly differentiates these cell types based on intrinsic motility, which provides a highly sensitive and efficient alternative to conventional particle tracing.