Biology subjects

Zarringhalam, K.

Publications and source records attributed to Zarringhalam, K..

16 recordsLinked to original sources

bioRxiv2026

A MultiOmics time-resolved systems map of gene regulation in Toxoplasma gondii

Understanding how T. gondii builds an invasion-competent cell has been limited by its rapid, asynchronous division cycle, incomplete gene function annotation, and studies that have examined one regulator or complex at a time. Here, we placed acute perturbation of 21 transcriptional and epigenetic factors, promoter occupancy, chromatin accessibility, and molecular validation on a common single-cell clock. The resulting consensus map links the lytic-cycle programs to candidate activating and repressive inputs, organizes genes into 108 co-regulated communities, and nominates regulatory partnerships that would not have emerged from any single data type. The work bridges three longstanding gaps: it resolves perturbation responses in cell-cycle time, distinguishes promoter occupancy from perturbation-linked regulation, and turns a large multidimensional dataset into an experimentally navigable map. We indeed confirmed known factor interactions such as MORC/HDAC3 and recovered combinations such as the Cdc5/SNIP1 and DMAP1/EPL1 as coregulatory nodes (that in other systems have been described to cooperate). Furthermore, we discovered a novel apicomplexan specific combination of AAF1/AP2XI-5 that provides mechanistic insights of the unique AP2 transcription factor family.

microbiology↗

bioRxiv2025

Single-cell analysis of brain-derived Toxoplasma bradyzoites reveals a novel cell cycle regulated by AP2XI-6.

Toxoplasma gondii prevalence is due, in part, to its ability to persist in hosts while retaining the capacity to transmit and recrudesce. A process that is poorly understood. Through single-cell RNA profiling of in vivo-derived bradyzoites, we discovered that they are heterogeneous and not G1-arrested, as expected from in vitro studies. Instead, they progress through two cell cycles that branch from a single G1. While in G1b, in vivo-derived bradyzoites express cyst wall proteins predicted to replenish the wall that shields cysts. One cell cycle is common amongst tachyzoites and in vitro- and in vivo-derived bradyzoites; the other is unique to in vivo-derived bradyzoites (BCC). We demonstrated that AP2XI-6, expressed in G1 in vivo, is dispensable for tachyzoite growth and cyst formation in vitro but necessary for encystation in vivo. We propose that AP2XI-6 is a driver of replication through the BCC that is required for cyst maturation and persistence.

microbiology↗

bioRxiv2025

Artificial intelligence-enabled automated analysis of transmission electron micrographs to evaluate chemotherapy impact on mitochondrial morphology in triple negative breast cancer

Advancements in transmission electron microscopy (TEM) have enabled in-depth studies of biological specimens, offering new avenues to large-scale imaging experiments with subcellular resolution. Mitochondrial structure is of growing interest in cancer biology due to its crucial role in regulating the multi-faceted functions of mitochondria. We and others have established the crucial role of mitochondria in triple-negative breast cancer (TNBC), an aggressive subtype of breast cancer with limited therapeutic options. Building upon our previous work demonstrating the functional role of mitochondrial structure dynamics in the metabolic adaptations and survival of chemotherapy-refractory TNBC cells, we sought to extend those findings to a large-scale analysis of transmission electron micrographs. Here we present a novel U-Net artificial intelligence (AI) model for automatic annotation and assessment of mitochondrial morphology and feature quantification. Our model is trained on 11,039 manually annotated mitochondria across 125 micrographs derived from a variety of orthotopic patient-derived xenograft (PDX) mouse model tumors and adherent cell cultures. The model achieves an F1 score of 0.85 on test micrographs at the pixel level. To validate the ability of our model to detect expected mitochondrial structural changes, we utilized micrographs from mouse primary skeletal muscle cells genetically modified to lack Dynamin-related protein 1 (Drp1). We subjected in vitro and in vivo TNBC models to conventional chemotherapy treatments commonly used for clinical management of TNBC, including doxorubicin, carboplatin, paclitaxel, and docetaxel (DTX). We found substantial within-sample heterogeneity of mitochondrial structure in both in vitro and in vivo TNBC models. In four of five PDX models, in vivo treatment with DTX elicited significant alteration in mitochondrial elongation and/or area. We went on to compare mammary tumors and matched lung metastases in a highly metastatic PDX model of TNBC, uncovering significant increase in mitochondrial elongation in metastatic lesions compared to their matched primary mammary tumor. The successful application of our AI model to capture mitochondrial structure marks a step forward in high-throughput analysis of mitochondrial structures, enhancing our understanding of how morphological changes may relate to chemotherapy efficacy and mechanism of action. Our large, manually curated electron micrograph dataset - now publicly available - serves as a unique resource for developing, benchmarking, and applying computational models, while further advancing investigations into mitochondrial morphology and its impact on breast cancer biology. This study provides proof of concept that mitochondrial structural remodeling is an additional layer of cellular reprogramming accompanying therapeutic resistance TNBC that merits further investigation.

cancer biology↗

bioRxiv2024

Drug resistant pancreatic cancer cells exhibit altered biophysical interactions with stromal fibroblasts in imaging studies of 3D co-culture models

Interactions between tumor and stromal cells are well known to play a prominent roles in progression of pancreatic ductal adenocarcinoma (PDAC). As knowledge of stromal crosstalk in PDAC has evolved, it has become clear that cancer associated fibroblasts can play both tumor promoting and tumor suppressive roles through a combination of paracrine crosstalk and juxtacrine interactions involving direct physical contact. Another major contributor to dismal survival statistics for PDAC is development of resistance to chemotherapy drugs. Though less is known about how the acquisition of chemoresistance impacts upon tumor-stromal crosstalk. Here, we use 3D co-culture geometries to recapitulate juxtacrine interactions between epithelial and stromal cells. In particular, extracellular matrix (ECM) overlay cultures in which stromal cells (pancreatic stellate cells, or normal human fibroblasts) are placed adjacent to PDAC cells (PANC1), result in direct heterotypic cell adhesions accompanied by dramatic fibroblast contractility which leads to highly condensed macroscopic multicellular aggregates as detected using particle image velocimetry (PIV) analysis to quantify cell velocities over the course of time lapse movie sequences. To investigate how drug resistance impacts these juxtacrine interactions we contrast cultures in which PANC1 are substituted with a drug resistant subline (PANC1-OR) previously established in our lab. We find that heterotypic cell-cell interactions are highly suppressed in drug-resistant cells relative to the parental PANC1 cells. To investigate further we conduct RNA-seq and bioinformatics analysis to identify differential gene expression in PANC1 and PANC1-OR, which shows that negative regulation of cell adhesion molecules, consistent with increased epithelial mesenchymal transition (EMT), is also consistent with loss of hetrotypic cell-cell contact necessary for the contractile behavior observed in drug naive cultures. Overall these findings elucidate the role of drug-resistance in inhibiting an avenue of stromal crosstalk which is associated with tumor suppression and also help to establish cell culture conditions useful for further mechanistic investigation.

biophysics↗

bioRxiv2023

Single cell expression and chromatin access of the Toxoplasma gondii lytic cycle identifies AP2XII-8 as an essential pivotal controller of a ribosome regulon

Sequential lytic cycles driven by cascading transcriptional waves underlie pathogenesis in the apicomplexan parasite Toxoplasma gondii. This parasites unique division by internal budding, short cell cycle, and jumbled up classically defined cell cycle stages have restrained in-depth transcriptional program analysis. Here, unbiased transcriptome and chromatin accessibility maps throughout the lytic cell cycle were established at the single cell level. Correlated pseudo-timeline assemblies of expression and chromatin profiles mapped transcriptional versus chromatin level transition points promoting the cell division cycle. Sequential clustering analysis identified putatively functionally related gene groups facilitating parasite division. Promoter DNA motif mapping revealed patterns of combinatorial regulation. Pseudo-time trajectory analysis revealed transcriptional bursts at different cell cycle points. The dominant burst in G1 was driven by transcription factor AP2XII-8, which engages TGCATGCG/A and TATAAGCCG motifs, and promoted the expression of a regulon encoding 40 ribosomal proteins. Overall, the study provides integrated, multi-level insights into apicomplexan transcriptional regulation.

microbiology↗

bioRxiv2023

Comparative chemical genomics in Babesia species identifies the alkaline phosphatase phoD as a novel determinant of resistance

Babesiosis is an emerging zoonosis and widely distributed veterinary infection caused by 100+ species of Babesia parasites. The diversity of Babesia parasites, coupled with the lack of potent inhibitors necessitates the discovery of novel conserved druggable targets for the generation of broadly effective antibabesials. Here, we describe a comparative chemogenomics (CCG) pipeline for the identification of novel and conserved targets. CCG relies on parallel in vitro evolution of resistance in independent populations of evolutionarily-related Babesia spp. (B. bovis and B. divergens). We identified a potent antibabesial inhibitor from the Malaria Box, MMV019266. We were able to select for resistance to this compound in two species of Babesia, achieving 10-fold or greater resistance after ten weeks of intermittent selection. After sequencing of multiple independently derived lines in the two species, we identified mutations in a single conserved gene in both species: a membrane-bound metallodependent phosphatase (putatively named PhoD). In both species, the mutations were found in the phoD-like phosphatase domain, proximal to the predicted ligand binding site. Using reverse genetics, we validated that mutations in PhoD confer resistance to MMV019266. We have also demonstrated that PhoD localizes to the endomembrane system and partially with the apicoplast. Finally, conditional knockdown and constitutive overexpression of PhoD alter the sensitivity to MMV019266 in the parasite: overexpression of PhoD results in increased sensitivity to the compound, while knockdown increases resistance, suggesting PhoD is a resistance mechanism. Together, we have generated a robust pipeline for identification of resistance loci, and identified PhoD as a novel determinant of resistance in Babesia species. HighlightsO_LIUse of two species for in vitro evolution identifies a high confidence locus associated with resistance C_LIO_LIResistance mutation in phoD was validated using reverse genetics in B. divergens C_LIO_LIPerturbation of phoD using function genetics results in changes in the level of resistance to MMV019266 C_LIO_LIEpitope tagging reveals localization to the ER/apicoplast, a conserved localization with a similar protein in diatoms C_LIO_LITogether, phoD is a novel resistance determinant in multiple Babesia spp. C_LI

molecular biology↗

bioRxiv2023

A Bayesian Noisy Logic Model for Inference of Transcription Factor Activity from Single Cell and Bulk Transcriptomic Data

The advent of high-throughput sequencing has made it possible to measure the expression of genes at relatively low cost. However, direct measurement of regulatory mechanisms, such as Transcription Factor (TF) activity is still not readily feasible in a high-throughput manner. Consequently, there is a need for computational approaches that can reliably estimate regulator activity from observable gene expression data. In this work, we present a noisy Boolean logic Bayesian model for TF activity inference from differential gene expression data and causal graphs. Our approach provides a flexible framework to incorporate biologically motivated TF-gene regulation logic models. Using simulations and controlled over-expression experiments in cell cultures, we demonstrate that our method can accurately identify TF activity. Moreover, we apply our method to bulk and single cell transcriptomics measurements to investigate transcriptional regulation of fibroblast phenotypic plasticity. Finally, to facilitate usage, we provide user-friendly software packages and a web-interface to query TF activity from user input differential gene expression data: https://umbibio.math.umb.edu/nlbayes/. Author SummaryNextGen RNA sequencing (RNA-Seq) has enabled simultaneous measurement of the expression level of all genes. Measurements can be done at the population level or single-cell resolution. However, direct measurement of regulatory mechanisms, such as Transcription Factor (TF) activity, is still not possible in a high-throughput manner. As such, there is a need for computational models to infer regulator activity from gene expression data. In this work, we introduce a Bayesian methodology that utilizes prior biological knowledge on bio-molecular interactions in conjunction with readily available gene expression measurements to estimate TF activity. The Bayesian model naturally incorporates biologically motivated combinatorial TF-gene interaction logic models and accounts for noise in gene expression data as well as prior knowledge. The method is accompanied by efficiently implemented R and Python software packages as well as a user-friendly web-based interface that allows users to upload their gene expression data and run queries on a TF-gene interaction network to identify and rank putative transcriptional regulators. This tool can be used for a wide range of applications, such as identification of TFs downstream of signaling events and environmental or molecular perturbations, the aberration in TF activity in diseases, and other studies with case-control gene expression data.

bioinformatics↗

bioRxiv2022

High-resolution small RNAs landscape provides insights into alkane adaptation in the marine alkane-degrader Alcanivorax dieselolei B-5

Alkanes are widespread in the ocean, and Alcanivorax is one of the most ubiquitous alkane-degrading bacteria in the marine ecosystem. Small RNAs (sRNAs) are usually at the heart of regulatory pathways, but sRNA-mediated alkane metabolic adaptability still remains largely unknown due to the difficulties of identification. Here, differential RNA sequencing (dRNA-seq) modified with a size selection ([~]50-nt to 500-nt) strategy was used to generate high-resolution sRNAs profiling in the model species Alcanivorax dieselolei B-5 under alkane (n-hexadecane) and non-alkane (acetate) conditions. As a result, we identified 549 sRNA candidates at single-nucleotide resolution of 5-ends, with 63.4% of transcription start sites (TSSs) and 36.6% of processing sites (PSSs). These sRNAs originated from almost any locations in the genome, regardless of intragenic (65.8%), antisense (20.6%) and intergenic (6.2%) regions, and RNase E may function in the maturation of sRNAs. Most sRNAs locally distribute across the 15 reference genomes of Alcanivorax, and only 7.5% of sRNAs are broadly conserved in this genus. Expression responses to alkane of several core conserved sRNAs, including 6S RNA, M1 RNA and tmRNA, indicate that they may participate in alkane metabolisms and result in more actively global transcription, RNA processing and stresses mitigation. Two novel CsrA-related sRNAs are identified, which may be involved in the translational activation of alkane metabolism-related genes by sequestering the global repressor CsrA. The relationships of sRNAs with characterized genes of the alkane sensing (ompS), chemotaxis (mcp, cheR, cheW2), transporting (ompT1, ompT2, ompT3) and hydroxylation (alkB1, alkB2, almA) were created based on the genome-wide targets prediction. Overall, the sRNAs landscape lays the ground for uncovering cryptic regulations in the critical marine bacterium, among which both core and species-specific sRNAs are implicated in the alkane adaptive metabolisms.

microbiology↗

bioRxiv2022

Babesia divergens egress from host cells is orchestrated by essential and druggable kinases and proteases

A unique aspect of apicomplexan biology is the requirement for egress from and invasion into host red blood cells (RBCs). The cellular mechanisms and molecular mediators of RBC egress and invasion remain poorly characterized in Babesia spp., a group of parasites of veterinary importance and emerging cause of zoonotic disease. Through the use of video microscopy, transcriptomics, and chemical genetics we have implicated signaling, proteases and gliding motility in egress and/or invasion by Babesia divergens. We developed CRISPR/Cas9 and two inducible knockdown systems to perform a genetic screen of putative mediators of egress. We found that proteases ASP2 and ASP3 are required for invasion, and the latter is also required for egress. Strikingly, parasites continue to replicate intracellularly in the absence of the protein kinases, PKG or CDPK4, indicating that they are required for exit from the replication cycle and egress. These essential molecules present druggable targets for Babesia spp. All together we have established a molecular framework for the spread of infection through host RBCs, with egress of B. divergens more closely resembling T. gondii than the more evolutionarily related Plasmodium spp. HighlightsO_LIEgress in Babesia divergens requires host cell lysis and parasite motility C_LIO_LITranscriptomics can be used to identify egress and invasion proteins C_LIO_LIKnockdown of the proteases, ASP2 and ASP3, inhibit egress and invasion C_LIO_LIInhibition of PKG or CDPK4 signaling results in continued intracellular replication C_LI

microbiology↗

bioRxiv2022

Comparative single-cell transcriptional atlases of Babesia species reveal conserved and species-specific expression profiles

Babesia is a genus of Apicomplexan parasites that infect red blood cells in vertebrate hosts. Pathology occurs during rapid replication cycles in the asexual blood-stage of infection. Current knowledge of Babesia replication cycle progression and regulation is limited and relies mostly on comparative studies with related parasites. Due to limitations in synchronizing Babesia parasites, fine-scale time-course transcriptomic resources are not readily available. Single-cell transcriptomics provides a powerful unbiased alternative for profiling asynchronous cell populations. Here, we applied single-cell RNA sequencing to three Babesia species (B. divergens, B. bovis, and B. bigemina). We used analytical approaches and algorithms to map the replication cycle and construct pseudo-synchronized time-course gene expression profiles. We identify clusters of co-expressed genes showing just-in-time expression profiles, with gradually cascading peaks throughout asexual development. Moreover, clustering analysis of reconstructed gene curves reveals coordinated timing of peak expression in epigenetic markers and transcription factors. Using a regularized Gaussian Graphical Model, we reconstructed co-expression networks and identified conserved and species-specific nodes. Motif analysis of a co-expression interactome of AP2 transcription factors identified specific motifs previously reported to play a role in DNA replication in Plasmodium species. Finally, we present an interactive web-application to visualize and interactively explore the datasets.

bioinformatics↗

bioRxiv2021

Deep learning trained on H&E tumor ROIs predicts HER2 status and Trastuzumab treatment response in HER2+ breast cancer

The current standard of care for many patients with HER2-positive breast cancer is neoadjuvant chemotherapy in combination with anti-HER2 agents, based on HER2 amplification as detected by in situ hybridization (ISH) or protein immunohistochemistry (IHC). However, hematoxylin & eosin (H&E) tumor stains are more commonly available, and accurate prediction of HER2 status and anti-HER2 treatment response from H&E would reduce costs and increase the speed of treatment selection. Computational algorithms for H&E have been effective in predicting a variety of cancer features and clinical outcomes, including moderate success in predicting HER2 status. In this work, we present a novel convolutional neural network (CNN) approach able to predict HER2 status with increased accuracy over prior methods. We trained a CNN classifier on 188 H&E whole slide images (WSIs) manually annotated for tumor regions of interest (ROIs) by our pathology team. Our classifier achieved an area under the curve (AUC) of 0.90 in cross-validation of slide-level HER2 status and 0.81 on an independent TCGA test set. Within slides, we observed strong agreement between pathologist annotated ROIs and blinded computational predictions of tumor regions / HER2 status. Moreover, we trained our classifier on pre-treatment samples from 187 HER2+ patients that subsequently received trastuzumab therapy. Our classifier achieved an AUC of 0.80 in a five-fold cross validation. Our work provides an H&E-based algorithm that can predict HER2 status and trastuzumab response in breast cancer at an accuracy that is better than IHC and may benefit clinical evaluations.

bioinformatics↗

bioRxiv2021

The extracellular milieu of Toxoplasma-s lytic cycle drives lab-adaptation and promotes changes in lipid metabolism primarily driven by transcriptional reprogramming

To map host-independent in vitro virulence traits of Toxoplasma gondii, evolve and resequencing (E&R) during the lab-adaption was applied. Phenotypic assessments of the lytic cycle revealed that only traits needed in the extracellular milieu evolved. Surprisingly, only non-synonymous mutations in a P4 flippase fixed in two populations. However, dramatic changes in the transcriptional signature of extracellular parasites revealed a "pro-tachyzoite" profile as well as upregulation of fatty acid biosynthesis (FASII) pathway genes. More general, a set of 300 genes which expression profile changes during evolution mapped to specific traits. Validation of a select number of genes in this set by knock-outs indeed confirmed their role in lab-adaptation. Finally, assembly of an ApiAP2 and Myb transcription factor network revealed the transcriptional program underlying the adapting extracellular state. Overall, E&R is a new genomic tool successfully applied to map the development of polygenic traits underlying in vitro virulence of T. gondii.

microbiology↗

bioRxiv2019

Pan-cancer classifications of tumor histological images using deep learning

Histopathological images are a rich but incompletely explored data type for studying cancer. Manual inspection is time consuming, making it challenging to use for image data mining. Here we show that convolutional neural networks (CNNs) can be systematically applied across cancer types, enabling comparisons to reveal shared spatial behaviors. We develop CNN architectures to analyze 27,815 hematoxylin and eosin slides from The Cancer Genome Atlas for tumor/normal, cancer subtype, and mutation classification. Our CNNs are able to classify tumor/normal status of whole slide images (WSIs) in 19 cancer types with consistently high AUCs (0.995{+/-}0.008), as well as subtypes with lower but significant accuracy (AUC 0.87{+/-}0.1). Remarkably, tumor/normal CNNs trained on one tissue are effective in others (AUC 0.88{+/-}0.11), with classifier relationships also recapitulating known adenocarcinoma, carcinoma, and developmental biology. Moreover, classifier comparisons reveal intra-slide spatial similarities, with average tile-level correlation of 0.45{+/-}0.16 between classifier pairs. Breast cancers, bladder cancers, and uterine cancers have spatial patterns that are particularly easy to detect, suggesting these cancers can be canonical types for image analysis. Patterns for TP53 mutations can also be detected, with WSI self- and cross-tissue AUCs ranging from 0.65-0.80. Finally, we comparatively evaluate CNNs on 170 breast and colon cancer images with pathologist-annotated nuclei, finding that both cellular and intercellular regions contribute to CNN accuracy. These results demonstrate the power of CNNs not only for histopathological classification, but also for cross-comparisons to reveal conserved spatial biology.

bioinformatics↗

bioRxiv2019

Causal Inference Engine: A platform for directional gene set enrichment analysis and inference of active transcriptional regulators

Inference of active regulatory mechanisms underlying specific molecular and environmental perturbations is essential for understanding cellular response. The success of inference algorithms relies on the quality and coverage of the underlying network of regulator-gene interactions. Several commercial platforms provide large and manually-curated regulatory networks and functionality to perform inference on these networks. Adaptation of such platforms for open-source academic applications has been hindered by the lack of availability of accurate, high-coverage networks of regulatory interactions and integration of efficient causal inference algorithms. In this work, we present CIE, an integrated platform for causal inference of active regulatory mechanisms form differential gene expression data. Using a regularized Gaussian Graphical Model, we construct a transcriptional regulatory network by integrating publicly available ChIP-Seq experiments with gene-expression data from tissue-specific RNA-Seq experiments. Our GGM approach identifies high confidence TF-gene interactions and annotates the interactions with information on mode of regulation (activation vs. repression). Benchmarks against manually-curated databases of TF-gene interactions show that our method can accurately detect mode of regulation. We demonstrate the ability of our platform to identify active transcriptional regulators by using controlled in vitro overexpression and stem-cell differentiation studies and utilize our method to investigate transcriptional mechanisms of fibroblast phenotypic plasticity.

bioinformatics↗

bioRxiv2019

ModEx: A text mining system for extracting mode of regulation of Transcription Factor-gene regulatory interaction

BackgroundTranscription factors (TFs) are proteins that are fundamental to transcription and regulation of gene expression. Each TF may regulate multiple genes and each gene may be regulated by multiple TFs. TFs can act as either activator or repressor of gene expression. This complex network of interactions between TFs and genes underlies many developmental and biological processes and is implicated in several human diseases such as cancer. Hence deciphering the network of TF-gene interactions with information on mode of regulation (activation vs. repression) is an important step toward understanding the regulatory pathways that underlie complex traits. There are many experimental, computational, and manually curated databases of TF-gene interactions. In particular, high-throughput ChIP-Seq datasets provide a large-scale map or transcriptional regulatory interactions. However, these interactions are not annotated with information on context and mode of regulation. Such information is crucial to gain a global picture of gene regulatory mechanisms and can aid in developing machine learning models for applications such as biomarker discovery, prediction of response to therapy, and precision medicine.\n\nMethodsIn this work, we introduce a text-mining system to annotate ChIP-Seq derived interaction with such meta data through mining PubMed articles. We evaluate the performance of our system using gold standard small scale manually curated databases.\n\nResultsOur results show that the method is able to accurately extract mode of regulation with F-score 0.77 on TRRUST curated interaction and F-score 0.96 on intersection of TRUSST and ChIP-network. We provide a HTTP REST API for our code to facilitate usage.\n\nAvailibilitySource code and datasets are available for download on GitHub: https://github.com/samanfrm/modex HTTP REST API: https://watson.math.umb.edu/modex/[type query]

bioinformatics↗

bioRxiv2019

Patch-DCA: Improved Protein Interface Prediction by utilizing Structural Information and Clustering DCA scores

Over the past decade there have been impressive advances in determining the 3D structures of protein complexes. However, there are still many complexes with unknown structures, even when the structures of the individual proteins are known. The advent of protein sequence information provides an opportunity to leverage evolutionary information to enhance the accuracy of protein-protein interface prediction. To this end, several statistical and machine learning methods have been proposed. In particular, direct coupling analysis has recently emerged as a promising approach for identification of protein contact maps from sequential information. However, the ability of these methods to detect protein-protein inter-residue contacts remains relatively limited.\n\nIn this work, we propose a method to integrate sequential and co-evolution information with structural and functional information to increase the performance of protein-protein interface prediction. Further, we present a post-processing clustering method that improves the average relative F1 score by 70 % and 24 % and the precision by 80 % and 36 % in comparison with two state-of-the-art methods PSICOV and GREMLIN.

bioinformatics↗