arXiv subjects

Roshan Prizak

Publications and source records attributed to Roshan Prizak.

5 recordsLinked to original sources

arXiv2023

AI-based automated active learning for discovery of hidden dynamic processes: A use case in light microscopy

In the biomedical environment, experiments assessing dynamic processes are primarily performed by a human acquisition supervisor. Contemporary implementations of such experiments frequently aim to acquire a maximum number of relevant events from sometimes several hundred parallel, non-synchronous processes. Since in some high-throughput experiments, only one or a few instances of a given process can be observed simultaneously, a strategy for planning and executing an efficient acquisition paradigm is essential. To address this problem, we present two new methods in this paper. The first method, Encoded Dynamic Process (EDP), is Artificial Intelligence (AI)-based and represents dynamic processes so as to allow prediction of pseudo-time values from single still images. Second, with Experiment Automation Pipeline for Dynamic Processes (EAPDP), we present a Machine Learning Operations (MLOps)-based pipeline that uses the extracted knowledge from EDP to efficiently schedule acquisition in biomedical experiments for dynamic processes in practice. In a first experiment, we show that the pre-trained State-Of-The- Art (SOTA) object segmentation method Contour Proposal Networks (CPN) works reliably as a module of EAPDP to extract the relevant object for EDP from the acquired three-dimensional image stack.

cs.CV↗

arXiv2018

Receptor crosstalk improves concentration sensing of multiple ligands

Cells need to reliably sense external ligand concentrations to achieve various biological functions such as chemotaxis or signaling. The molecular recognition of ligands by surface receptors is degenerate in many systems leading to crosstalk between different receptors. Crosstalk is often thought of as a deviation from optimal specific recognition, as the binding of non-cognate ligands can interfere with the detection of the receptor's cognate ligand, possibly leading to a false triggering of a downstream signaling pathway. Here we quantify the optimal precision of sensing the concentrations of multiple ligands by a collection of promiscuous receptors. We demonstrate that crosstalk can improve precision in concentration sensing and discrimination tasks. To achieve superior precision, the additional information about ligand concentrations contained in short binding events of the non-cognate ligand should be exploited. We present a proofreading scheme to realize an approximate estimation of multiple ligand concentrations that reaches a precision close to the derived optimal bounds. Our results help rationalize the observed ubiquity of receptor crosstalk in molecular sensing.

q-bio.SC↗

arXiv2017

Evolution of new regulatory functions on biophysically realistic fitness landscapes

Regulatory networks consist of interacting molecules with a high degree of mutual chemical specificity. How can these molecules evolve when their function depends on maintenance of interactions with cognate partners and simultaneous avoidance of deleterious "crosstalk" with non-cognate molecules? Although physical models of molecular interactions provide a framework in which co-evolution of network components can be analyzed, most theoretical studies have focused on the evolution of individual alleles, neglecting the network. In contrast, we study the elementary step in the evolution of gene regulatory networks: duplication of a transcription factor followed by selection for TFs to specialize their inputs as well as the regulation of their downstream genes. We show how to coarse grain the complete, biophysically realistic genotype-phenotype map for this process into macroscopic functional outcomes and quantify the probability of attaining each. We determine which evolutionary and biophysical parameters bias evolutionary trajectories towards fast emergence of new functions and show that this can be greatly facilitated by the availability of "promiscuity-promoting" mutations that affect TF specificity.

q-bio.PE↗

arXiv2016

Intrinsic limits to gene regulation by global crosstalk

Gene regulation relies on the specificity of transcription factor (TF) - DNA interactions. In equilibrium, limited specificity may lead to crosstalk: a regulatory state in which a gene is either incorrectly activated due to noncognate TF-DNA interactions or remains erroneously inactive. We present a tractable biophysical model of global crosstalk, where many genes are simultaneously regulated by many TFs. We show that in the simplest regulatory scenario, a lower bound on crosstalk severity can be analytically derived solely from the number of (co)regulated genes and a suitable parameter that describes binding site similarity. Estimates show that crosstalk could present a significant challenge for organisms with low-specificity TFs, such as metazoans, unless they use appropriate regulation schemes. Strong cooperativity substantially decreases crosstalk, while joint regulation by activators and repressors, surprisingly, does not; moreover, certain microscopic details about promoter architecture emerge as globally important determinants of crosstalk strength. Our results suggest that crosstalk imposes a new type of global constraint on the functioning and evolution of regulatory networks, which is qualitatively distinct from the known constraints acting at the level of individual gene regulatory elements.

q-bio.MN↗

arXiv2010

Opinion formation and cyclic dominance in adaptive networks

The Rock-Paper-Scissors(RPS) game is a paradigmatic model for cyclic dominance in biological systems. Here we consider this game in the social context of competition between opinions in a networked society. In our model, every agent has an opinion which is drawn from the three choices: rock, paper or scissors. In every timestep a link is selected randomly and the game is played between the nodes connected by the link. The loser either adopts the opinion of the winner or rewires the link. These rules define an adaptive network on which the agent's opinions coevolve with the network topology of social contacts. We show analytically and numerically that nonequilibrium phase transitions occur as a function of the rewiring strength. The transitions separate four distinct phases which differ in the observed dynamics of opinions and topology. In particular, there is one phase where the population settles to an arbitrary consensus opinion. We present a detailed analysis of the corresponding transitions revealing an apparently paradoxial behavior. The system approaches consensus states where they are unstable, whereas other dynamics prevail when the consensus states are stable.

nlin.AO↗