Biology subjects

He, Z.

Publications and source records attributed to He, Z..

At least 55 records · Page 3Linked to original sources

bioRxiv2026

Single-crypt evolutionary trajectories of human sporadic and hereditary colorectal precancers

Our understanding of the earliest genomic processes and evolutionary trajectories that distinguish sporadic from hereditary human cancers remains limited, impeding early diagnosis and risk stratification. Here, we performed whole-genome sequencing of 319 single crypts dissected from normal colorectal mucosa, premalignant polyps and malignant colorectal cancers (CRCs) from five sporadic patients and three familial adenomatous polyposis (FAP) patients. We found 66% of premalignant crypts from sporadic polyps did not harbor APC mutations, but frequently accumulated non-canonical but positively selected driver mutations including KMT2C, CACNA1A and FBLN2. In contrast, most premalignant crypts (71%) in FAP indeed acquired one or more somatic APC mutations or 5q loss of heterogeneity (LOH). Sporadic polyps were dominated by clock-like mutational processes at single-crypt level. In contrast, FAP crypts evolved more diverse genomics processes, including reactive oxygen species-associated damages, APOBEC-associated mutagenesis and pks+ Escherichia coli-associated mutagenesis. Premalignant crypts in FAP also exhibited a 2-fold higher burden of LINE-1 retrotranspositions than sporadic ones, despite arising at a substantially younger age. Phylogenetic timing estimations showed that the first somatic APC mutation in FAP often occurs prenatally. Phylogeographic analysis further revealed spatial lineage segregation in early polyclonal polyps, followed by Big-Bang lineage intermixing in later monoclonal ones. Together, these data demonstrate that early colorectal tumorigenesis does not obey a universal APC-first model, where sporadic and hereditary precancers follow distinct evolutionary trajectories. Our study also highlights that polyclonal-to-monoclonal transition in premalignant stages represents the Big-Bang growth of colorectal tumors, offering crucial insight into early diagnosis and risk stratification.

cancer biology↗

bioRxiv2026

Integrative Identification and Characterization of PCOS-Associated lncRNAs From the Interface of Genetic Association, Transcriptomics, and Gene Structure Evolution

BackgroundPolycystic ovary syndrome (PCOS) is a prevalent endocrine disorder and a leading cause of female infertility, with complex genetic, metabolic, and hormonal etiologies. Long non-coding RNAs (lncRNAs) have emerged as important regulators of diverse biological processes, yet their roles in PCOS remain underexplored. Here, we identified and characterized PCOS differentially expressed gene-associated lncRNAs (PDEGAL) with an integrative approach combining expression data, genetic association, and evolutionary analysis. MethodsThirty-three PCOS-associated protein-coding genes were obtained from our prior study, and all their nearby and overlapping lncRNAs were annotated. These candidates were analyzed using UCSC Genome Browser-mapped annotations and datasets, including NCBI RefSeq, GENCODE, GTEx, GWAS SNPs, and conservation, as well as the FANTOM5 cap analysis of gene expression (CAGE) promoter data, to assess their expression, regulatory potential, genetic variant overlaps, and evolutionary conservation. ResultsTwenty-three PDEGALs (18 antisense to, and 5 sharing bidirectional promoters with, known PCOS-associated protein-coding genes) were identified. 17 PDEGALs contained GWAS SNPs with statistically significant disease associations, 9 of which were associated with PCOS-related traits. 5 PDEGALs demonstrated expression in the KGN granulosa cell model of PCOS. Key gene structure element (KGSE) analysis revealed that most PDEGALs are primate-specific. Integrating four criteria--GTEx expression, GWAS SNPs, FANTOM promoterome, and KGSE conservation--highlighted HELLPAR as the only lncRNA fulfilling all four, while five others--PGR-AS1, MTOR-AS1, ENSG00000265179, ENSG00000256218, and LOC105377276--fulfilled three of the four criteria. ConclusionsWe have systematically identified candidate PCOS regulatory lncRNAs with convergent genetic, expression, and evolutionary evidence. These results provide a framework for functional validation and highlight lncRNAs as potential biomarkers and therapeutic targets in PCOS that function by regulating their nearby and overlapping protein-coding genes.

genomics↗

bioRxiv2026

scMagnifier: resolving fine-grained cell subtypes via GRN-informed perturbations and consensus clustering

Resolving fine-grained cell subtypes in single-cell RNA sequencing (scRNA-seq) data remains challenging, as their subtle transcriptional differences are often obscured by technical noise and data sparsity. Here, we present scMagnifier, a consensus clustering framework that leverages gene regulatory network (GRN)-informed in silico perturbations to amplify subtle transcriptional differences and uncover latent cell subpopulations. scMagnifier perturbs candidate transcription factors (TFs), propagates perturbation effects through cluster-specific GRNs to simulate post-perturbation expression profiles, and integrates clustering results across multiple perturbations into stable subtype assignments. Additionally, scMagnifier introduces regulatory perturbation consensus UMAP (rpcUMAP), a perturbation-aware visualization that provides clearer separation between cell subtypes and guides the selection of the optimal number of clusters. In both single-batch and multi-batch benchmarks, scMagnifier consistently improves the resolution and accuracy of fine-grained cell type identification. Notably, when integrated with spatial clustering methods such as STAGATE, scMagnifier is compatible with spatial transcriptomics workflows and effectively reveals tumor cell subtypes and their spatial organization in ovarian cancer.

bioinformatics↗

bioRxiv2026

Mechanochemically-reprogrammed stem cell exosomes reconcile the biogenesis internalization trade-off for pulmonary fibrosis therapy

Matrix stiffness serves as a pivotal biophysical cue that profoundly dictates exosome biogenesis and cellular internalization, yet often creates a functional trade-off that impedes clinical translation. Herein, we developed a mechano-chemo-transductive strategy to engineer mesenchymal stem cell (MSC) exosomes endowed with robust biogenesis and superior delivery potency. Specifically, we revealed that MSCs cultured on soft matrices secreted a significantly elevated exosome yield and demonstrated enhanced competence to drive macrophage towards anti-inflammatory M2 polarization. Conversely, stiff matrices upregulated ATP-binding cassette transporter A1 (ABCA1) expression, enriching exosomal membrane cholesterol and facilitating cellular internalization by recipient cells. By taking advantages of these unique mechano-responses, we engineered MSCs via substrate softening combined with ABCA1 modulation to generate mechanochemically reprogrammed exosomes with concurrently enhanced yield and internalization efficiency. In a murine model of pulmonary fibrosis characterized by restrictive biological barriers, inhaled mechanochemically reprogrammed exosomes treatment demonstrated superior lung retention and deep tissue penetration. Furthermore, they effectively orchestrated immune homeostasis by repolarizing alveolar macrophages to reverse fibrotic remodeling and restore lung function. Collectively, by reconciling the intrinsic trade-off between biogenesis and cellular uptake, this strategy represents a paradigm shift in exosome engineering and paves the way for next-generation therapeutics against refractory fibrotic diseases.

bioengineering↗

bioRxiv2026

HMA proteins produce 2',3'-cNMP signaling molecules and activate CNL-mediated immunity in rice

Plants deploy a sensor repertoire, usually NLRs, to perceive pathogen effectors and activate executor NLR-triggered immunity. Here we reveal that rice HMA proteins function as sensors detecting the Magnaporthe oryzae MAX effector AvrPigm to trigger broad-spectrum blast resistance mediated by the CNL PigmR. AvrPigm is conserved with multiple copies in blast genomes and targets HMA proteins, facilitating HMA translocation into the cytoplasm. Cryo-EM structure of the HPP04HMA reveals a filament-like oligomer with ssDNA/RNA bound inside the filament, which display cyclic nucleotide synthase activity to generate 2',3'-cNMP in a CNL-dependent manner. We further discovered that the LRR domains of the executor CNLs perceive 2',3'-cNMP to mount immunity. Our study thus establishes an unrecognized immunity mode with a distinct repertoire of sensor receptors that produce signaling molecules that are perceived by the LRR domains of executor NLRs to mediate immunity, shedding new light on the sensor-executor conception and immune activation in plants. HighlightsO_LIHMA proteins and CNL receptors form a new sensor-executor mode in plant immunity C_LIO_LICryo-EM reveals the filament-like structure of HMA-ssRNA/ssDNA C_LIO_LIHMA proteins are cyclic nucleotide synthetases to generate 2',3'-cNMP C_LIO_LILRR domains of CNLs perceive 2',3'-cNMP to mediate immunity C_LI

plant biology↗

bioRxiv2026

Inducible activation of PKA in osteoblasts causes a profound high bone turnover phenotype similar to human diseases

Protein kinase A (PKA) is involved in bone biology and is a key mediator of parathyroid hormone signaling in the osteoblast. However, the consequences of sustained PKA activation in bone are unclear. In this study, we inducibly activated PKA in osteoblasts by deleting its major regulatory subunit, Prkar1a, using a Col11-driven Cre system. Prkar1aob-/-mice demonstrated rapid and profound bone pathologies in their femurs, lumbar and caudal vertebrae with cortical bone breakdown and cortical trabecularization. This phenotype was characterized by increased bone turnover and elevated osteoblastic and osteoclastic activities. Transcriptomic and qPCR analyses showed an impairment of osteoblast differentiation with a defect in ossification, expansion of stromal cells, and numbers of both osteoblastic and osteoclastic precursors. Moreover, there were alterations in gene expression of chemokines and Wnt members with enhanced osteoclastogenesis. Altogether, activation of PKA in osteoblasts by inducible deletion of Prkar1a causes a profound high bone turnover phenotype resembling several human bone diseases.

pathology↗

bioRxiv2026

Comparing Electrical and Ultrasound Transcutaneous Vagus Nerve Stimulation (taVNS) on Associative Memory

Associative memory, the ability to bind and retrieve relationships between unrelated elements, is a cornerstone of human cognition and a primary target for neurorehabilitation. Vagus nerve stimulation (VNS) has emerged as a promising method to modulate the locus coeruleus-norepinephrine (LC-NE) system and hippocampal-prefrontal circuits essential for memory. However, the comparative efficacy of non-invasive modalities such as electrical (E-taVNS) and the emerging field of ultrasound (U-taVNS) remains poorly understood in the context of active recall. In this study, participants performed a crossmodal video-word associative memory task before and after receiving either E-taVNS or U-taVNS in active and sham conditions. We investigated whether these modalities enhance cued recall accuracy and retrieval reaction time. Our results revealed that neither E-taVNS nor U-taVNS significantly improved recall accuracy. However, E-taVNS significantly accelerated response times specifically for correctly recalled items. These findings suggest that while taVNS may not increase the likelihood of recalling associative memories, electrical stimulation may enhances the efficiency in which we do so. These findings suggest that electrical taVNS is a viable tool for facilitating memory search processes, though further research is required to optimize ultrasound parameters and validate mechanistic pathways through physiological monitoring.

neuroscience↗

bioRxiv2026

Benchmarking foundation models for splice site and exon annotation

Recent foundation and deep learning models have brought a generational leap in improving the quality of genome annotation, particularly in identifying genes and their structural elements, including exons and splice sites. However, they are trained on reduced datasets that may not capture biological complexity, such as differences between coding versus non-coding, terminal versus internal, constitutive versus alternatively spliced, and transposable element (TE)-derived exons. We evaluate several foundation models for gene and splice site annotation, including the transformer-based SegmentNT, Enformer and Borzoi, coupled with a segmentation head for per-base resolution, and the CNN-based SpliceAI and AlphaGenome, along with a newly developed fine-tuned model, STEP2h, on different classes of gene elements as described above. We found that the performance of all methods is highest for the class of exons found in their training data class and decreases drastically for classes of exons poorly represented. In particular, performance is highest for protein-coding genes, coding exons, and constitutive exons, and decreases drastically by up to 2-4 fold for non-coding internal exons, terminal exons, and exons that undergo alternative splicing. Similarly, performance is impaired on LINE-1 and Alu-derived exons. In contrast, a locally developed CNN model fine-tuned on a specialized TE-exon dataset showed improved performance in this category. Our study highlights the outstanding challenges in gene and exon annotation when leveraging powerful foundation models, and the need for further fine-tuning on judiciously selected classes of data or task-specific models to capture a broader, more diverse spectrum of gene features.

bioinformatics↗

bioRxiv2026

TRIM32-UBQLN2-p62 axis promotes TDP-43 inclusion formation and amyloid aggregation through shuttle condensates

Aberrant protein aggregation is a hallmark of amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD), which share overlapping genetic and pathological features. Similar aggregates are increasingly recognized in Alzheimers disease (AD) and limbic-predominant age-related TDP-43 encephalopathy (LATE). However, it remains unclear whether a shared molecular pathway drives this pathological aggregation. Here, we report that the E3 ubiquitin ligase TRIM32, together with the shuttle factor UBQLN2 and the autophagy adaptor p62/SQSTM1, form condensates that depend on E3 ligase activity and a network of intermolecular interactions. These condensates act as scaffolds that capture UBQLN2 client proteins, including TDP-43 and ANXA11, and modulate their mobility. A unique hydrophobic loop within TRIM32s substrate-binding domain mimics low-complexity motifs in ANXA11 and TDP-43, enabling selective retention via competitive binding mediated by UBQLN2 STI1 domain. Moreover, TRIM32 condensates promote amyloid aggregation of TDP-43, an effect that is exacerbated by pathogenic UBQLN2 mutation. In brains from individuals with diverse neurodegenerative diseases, TRIM32 co-localizes with pathological phospho-TDP-43 (pTDP-43) inclusions, supporting a model in which TRIM32-driven condensates function as selective proteostasis sorting compartments that broadly contribute to TDP-43 proteinopathy.

neuroscience↗

bioRxiv2026

Hierarchy in neuronal representations of multiple tasks in prefrontal cortex

The ability to perform multiple tasks is one of the fundamental hallmarks of general intelligence in the brain. However, the underlying neural mechanisms remain largely unknown. We trained macaque monkeys to perform four tasks with identical spatial layouts but different cognitive demands, and recorded activities of thousands of prefrontal neurons using two-photon calcium imaging. Within the same neuronal population, we identified a multitask neural geometry composed of multiple low-dimensional subtask spaces, each encoding spatial information within a specific subtask. These subtask spaces shared a ring-like representational structure, forming a generalized spatial code across different subtasks. Task separation arose from separable bases and offsets of subtask spaces, modulated primarily by meta-task features. Thus, the neural geometry is organized by a three-level hierarchical structure: location codes are nested within subtask spaces, which are grouped as different meta-tasks. This hierarchy supported generalization across both locations and tasks, and explained monkeys error behaviors. Together, the hierarchically organized neural geometry underlies flexible multitask behavior.

neuroscience↗

bioRxiv2026

Divergent Electrophysiological Responses in the Human Hippocampus During Verbal Memory Processing

The profile of electrophysiological responses in the human hippocampus (HPC) during verbal memory processing has remained complex and unclear. Here, we studied 26 patients implanted with intracranial electrodes across 187 HPC sites (50% left, 2-18 per patient). During memory encoding and retrieval, a subset of HPC responsive sites demonstrated increased ripple events, along with elevated high-frequency (HFA >50 Hz), and low-frequency (LFA 1-8 Hz) activity. A nearly equal number of sites showed no changes in ripple rate but increased LFA power and a delayed response-locked decrease in HFA power. More importantly, both successful encoding as well as recognition of remembered words were strongly associated with the coordination of the timing of LFA and HFA increases across the two clusters of responsive HPC sites. Using direct cortical electrical stimulations, we confirmed overlapping, but partially distinct, cortical connections to the functionally distinct HPC clusters. Our findings suggest a mesoscale mosaic functional organization within the human HPC where adjacent sites with divergent electrophysiological responses may have specialized roles during verbal memory processing. More importantly, our findings suggest that successful human memory depends on the coordination of the timing of low and high frequency local fields generated across these functionally divergent neuronal population sites.

neuroscience↗

bioRxiv2026

CCycDB: an integrative knowledgebase to fingerprint microbially mediated carbon cycling processes

Microorganisms play essential roles in mediating biogeochemical cycling of carbon across Earths ecosystems. Understanding the processes and underlying mechanisms for microbially mediated carbon cycling is therefore critical for advancing global ecology and climate change research. To comprehensively depict these complex biogeochemical processes, we developed CCycDB, a knowledge-based functional gene database, to accurately fingerprint microbially-mediated carbon cycling pathways and gene families, particularly from shotgun metagenomes. The CCycDB database comprises 4,676 gene families classified into six major functional categories, further structured into 45 level-1 and 188 level-2 sub-categories, encompassing a total of 10,991,724 high-quality reference sequences. Validation using both synthetic and real-world datasets demonstrated that CCycDB outperforms existing orthology databases in terms of accuracy, coverage and specificity. By directly targeting carbon-cycling functional gene families, CCycDB provided promising routines to reconstruct both functional gene and taxonomic profiles associated with microbially mediated carbon cycling. Application of CCycDB to shotgun metagenomes from diverse and complex ecosystems revealed pronounced habitat-specific differences in carbon cycling processes and their associated microbial taxa. Collectively, CCycDB provides a powerful and reliable tool for profiling carbon cycling processes from both functional and taxonomic perspectives in complex ecosystems. CCycDB is accessible at https://ccycdb.github.io/. Impact StatementThe microbially mediated carbon cycling processes are the most complex biogeochemical processes in the Earths biosphere, playing profound regulatory roles on global climate changes. A key bottleneck in linking microbial communities to global change is the lack of integrated tools for comprehensive carbon cycle profiling. Here, we present CCycDB, a tool that serves a dual purpose--first being a reference database that obtains functional gene and taxonomic profiles and functioning as a customized routine for efficiently aligning sequences and querying associated functional information. CCycDB enables researchers to accurately link microbial community dynamics to carbon cycling and transforming pathways, thereby advancing integrated global change studies with microbes and ecological research via complex metagenomic datasets.

microbiology↗

bioRxiv2026

Critical role of cell competition in gliomagenesis

ABSTRACTMalignant glioma is incurable. Using a mouse genetic mosaic system to generate sporadic Trp53,Nf1-null OPCs, we previously identified oligodendrocyte precursor cell (OPC) as a cell-of-origin of glioma. Here, we report that pre-malignant Trp53,Nf1-null OPCs outcompete wildtype counterparts during their expansion. Blocking competition by mutating/strengthening wildtype OPCs impeded both pre-malignant progression and malignant expansion of glioma. "In-tissue" phosphoproteomic profiling revealed an enrichment of phosphopeptides related to RNA splicing and protein translation at the peak of cell competition, suggesting that competitiveness may stem from unique protein species. Among candidates was mTORC1, whose pharmacological inhibition or genetic disruption resulted in a loss of competitiveness in our mouse model. Finally, analysis of patient biopsies and interrogating the role of individual gliomagenic mutations in OPC competition supported its relevance in human gliomas. Together, these findings identified the driving role of competitive interactions among OPCs in gliomagenesis, and suggest unconventional therapeutic strategies to target this process.

cancer biology↗

bioRxiv2025

Semi-rational Engineering of Terminal Deoxynucleotidyl Transferase for High-efficiency Enzymatic de Novo DNA Synthesis

The 21st century has witnessed rapid advancements in synthetic biology, with DNA synthesis emerging as a foundational technology. Conventional phosphoramidite-based methods face significant limitations, including short DNA elongation lengths (<300 nt), hazardous chemical waste, and low stepwise incorporation efficiency. Enzymatic DNA synthesis using terminal deoxynucleotidyl transferase (TdT) offers a promising alternative, enabling kilobase-scale assembly with greater efficiency and minimal environmental impact. Here, we identified Bos taurus TdT (BtTdT) through UniProt database mining as a catalytically active scaffold for natural and 3'-modified dNTPs. Comprehensive characterization of BtTdTs enzymatic properties--including pH, temperature, metal ion dependence, and substrate specificity--revealed its optimal conditions. Truncation of the BRCT domain generated variants with enhanced activity compared to wild-type BtTdT. Guided by AlphaFold3-predicted structural models, we engineered a quintuple mutant (M5: Bt15AAR336L/K338G/L397M/E456S/D395G) optimized for 3'-ONH2-dNTP incorporation. M5 exhibited 30-fold activity enhancement relative to the triple mutant M3 (Bt15AAR336L/K338G/L397M) and achieved stepwise incorporation efficiency exceeding 98% in de novo synthesis of 10-nt ssDNA, demonstrating its potential for scalable enzymatic DNA synthesis. This work establishes a rational framework for TdT engineering through rational domain truncation and computational design, showing potential toward industrial-scale enzymatic DNA manufacturing.

synthetic biology↗

bioRxiv2025

PocketX: Preference Alignment for Protein Pockets Design through Group Relative Policy Optimization

Designing protein pockets that target specific ligands is crucial for drug discovery and enzyme engineering. Although deep generative models show promise in proposing high-quality pockets, they are usually trained purely to match the data distribution and therefore overlook key biophysical properties, such as binding affinity, expression, and solubility, that ultimately determine developability and success. We introduce PocketX, an online reinforcement learning framework that explicitly aligns a generative model with desired biophysical properties. The framework first trains a base model that co-designs pocket structures and sequences conditioned on a target ligand, and then fine-tunes this model with Group Relative Policy Optimization (GRPO) to reward the desired attributes. Because GRPO employs group-relative rewards, it produces lower-variance policy updates, resulting in more stable and efficient learning than competing alignment strategies. Evaluated on the CrossDocked2020 benchmark, PocketX surpasses existing methods in metrics such as binding energy and evolutionary plausibility. Ablation studies further show that GRPO outperforms alternative alignment strategies, including Direct Preference Optimization (DPO), confirming GRPOs effectiveness for biophysical property alignment.

bioinformatics↗

bioRxiv2025

Combinatorial Protein Language Model-Guided Engineering of TEV Protease for Enhanced Stability and Production

Industrial enzyme engineering focuses on improvement of enzyme production yield, stability, catalytic activity, and substrate specificity, but often suffers from low efficiency with time-consuming and labor-intensive design and screening processes of massive libraries. Recent advances in AI and machine learning created protein language models trained by numerous datasets and shed new lights to speed up the enzyme engineering processes with high accuracy structural prediction. Here, we developed a highly efficient enzyme engineering strategy combining three protein language models (xTrimoMPNN-Thermo, ESM-IF, and MPNNsol) and use it to generate TEV protease variants with improved expression, stability, and function. The results indicated that a small number of TEV protease designs (<50 designs) were sufficient to develop variants with desired properties, demonstrating its high efficiency. Our strategy could be broadly applied to accelerate designing and engineering various industrial enzymes.

bioengineering↗

bioRxiv2025

Direction and orientation preferences in mouse superior colliculus and its retinal inputs exhibit a topography of cardinal biases atop locally mixed tuning

In mouse superior colliculus (SC), neurons are robustly tuned to motion direction and orientation, but it remains controversial whether these preferences are spatially organized, and how this organization relates to that found in the retinal input to the SC. We addressed these questions by combining two-photon calcium imaging of retinal boutons and SC neurons in superficial SC with Neuropixels recordings across the full depth of SC. We then asked how direction and orientation preferences depend on visual-field location and how tuning similarity among units depends on lateral and vertical distance in the SC. Retinal boutons were strongly tuned, and their preferences closely matched the retinal topographic organization of four motion directions and orientations previously described in the retina, rather than a global map in which a single direction or orientation dominates each visual-field location. With increasing depth, SC neurons progressively deviated from this organization. Superimposed on the identified topography, local clustering of tuning preferences in retinal boutons and SC neurons was weak and confined to very small spatial scales. Together, these results show that retinal inputs and neurons in mouse SC represent multiple directions and orientations for each location in the visual field, likely supporting flexible readout for diverse visually guided behaviors.

neuroscience↗

bioRxiv2025

SPOmiAlign: A modality-agnostic framework for robust and scalable spatial multimodal alignment via feature matching

Multimodal spatial omics enables systematic characterization of tissue organization by jointly profiling molecular features across transcriptomic, proteomic, and metabolomic layers. However, integrative analysis across sections and modalities is frequently hindered by non-linear tissue distortions, mismatched spatial resolutions, and the absence of shared molecular features. To address these challenges, we introduce SPOmiAlign, a modality-agnostic framework based on uncertainty-aware dense feature matching for spatial multimodal alignment. The framework achieves second-level runtime efficiency and enables accurate alignment across heterogeneous spatial omics datasets, as well as between spatial omics and histology, without manual effort or modality-specific tuning. Across diverse datasets, SPOmiAlign consistently achieves higher alignment accuracy than existing state-of-the-art methods. We further demonstrate its utility through automated registration of spatial omics data to a common coordinate framework (CCF), enabling anatomical annotation. Applying SPOmiAlign to spatial multi-omic data demonstrates how accurate multimodal alignment is essential for integrative spatial multi-omic analysis, enabling the identification of coherent spatial domains across modalities and supporting a broad range of downstream spatial analysis.

bioinformatics↗