Structural Annotation Research Articles

Real-time telementoring with 3D drawing annotation in robotic surgery.

In telementoring, differences in teaching methods affect local surgeons' comprehension. Because the object to be operated on is a three-dimensional (3D) structure, voice or 2D annotation may not be sufficient to convey the instructor's intention. In this study, we examined the usefulness of telementoring using 3D drawing annotations in robotic surgery. Kyushu University and Beppu Hospital are located 140km apart, and the study was conducted using a Saroa™ surgical robot by RIVERFIELD Inc. using a commercial guarantee network on optical fiber. Twenty medical students performed vertical mattress suturing using a swine intestinal tract under surgical guidance at the Center for Advanced Medical Innovation Kyushu University. Surgical guidance was provided by Beppu Hospital using voice, 2D, and 3D drawing annotations. All robot operations were performed using 3D images, and only the annotations were independently switched between voice and 2D and 3D images. The operation time, needle movement, and performance were also evaluated. The 3D annotation group tended to have a shorter working time than the control group (25.6 ± 63.2 vs. -36.7 ± 65.4min, P = 0.06). The 3D annotation group had fewer retries than the control group (1.3 ± 1.7 vs. -1.1 ± 0.7, P = 0.006), and there was a tendency for fewer needle drops (0.4 ± 0.7 vs. -0.5 ± 0.9, P = 0.06). The 3D annotation group scored significantly higher than the control group on the Global Evaluate Assessment of Robot Skills (16.8 ± 2.0 vs. 22.8 ± 2.4, P = 0.04). The 3D annotation group also scored higher than the voice (13.4 ± 1.2) and 2D annotation (16.2 ± 1.8) groups (3D vs. voice: P = 0.03, 3D vs. 2D: P = 0.03). Telementoring using 3D drawing annotation was shown to provide good comprehension and a smooth operation for local surgeons.

NeoDescriber: An image-to-text model for automatic style description of neoclassical architecture

Architectural style description is important for the renewal and reuse of historical monuments. However, the identification of Neoclassical buildings is more difficult than that of other European architectural styles owing to the revival of classicism and its numerous variants. This paper proposes the image-to-text model NeoDescriber for the automatic identification and description of Neoclassical buildings based on the combination of dedicated visual classification and detection modules organized by expert knowledge in architecture. NeoDescriber takes building façade images as input and follows a coarse-to-fine description logic to generate structured annotations covering the whole building and key building elements. We built several building image datasets composed of more than 300 historical monuments to train and validate different visual modules. The results showed the acceptance performance of both classification and detection modules, with average precisions ranging from 73.6% to 95.2%. The generated text descriptions were then compared with results from BLIP-2 by qualitative and quantitative analyses. With human descriptions as a reference, the results showed that NeoDescriber’s generated descriptions were more precise and detailed, having a higher BERTScore for Neoclassical monuments. Thus, NeoDescriber could successfully recognize and describe Neoclassical buildings.

Mass spectrometry-based untargeted metabolomics approaches for comprehensive structural annotation of bioactive metabolites from bushy cashew (Anacardium humile) fruits

Anacardium humile (bushy cashew) is a native Brazilian plant with substantial pharmacological potential and noteworthy commercial significance in the food industry. This study introduces an untargeted metabolomics approach based on mass spectrometry for the comprehensive structural annotation of bioactive metabolites. The fruits were collected from three distinct sites and subjected to LC-HRMS/MS analysis. A total of eighty-eight compounds were putatively annotated across various metabolite classes. This unveiled a metabolic profile characterized by notable concentrations of polyphenols, including flavonoids, tannins, phenolic acids, and quinones, while aliphatic acids and terpenes were found in limited quantities. Noteworthy, no significant disparities in the metabolic content were observed among the collection sites. The three principal metabolites (peonidin 3-O-glucopyranoside, methylcyanidin and methyldelphinidin) underwent assessment for antioxidant activity via molecular docking analysis, subsequently generating structure-based feature pharmacophores. The process of annotation propagation yielded a comprehensive qualitative appraisal of the A. humile metabolome. The outcomes obtained offer potential candidates for further exploration of their nutraceutical attributes.

Genome-Wide Association and Pathway Analysis of Carcass and Meat Quality Traits in Karachai Young Goats.

Goats with diverse economic phenotypic traits play an important role in animal husbandry. However, the genetic mechanisms underlying complex phenotypic traits are unclear in goats. Genomic studies of variations provided a lens to identify functional genes. The work aimed to search for candidate genes related to body measurements and body weight of Karachai goats and develop an experimental PCR-RV test system for genotyping significant SNPs. Comparison of GWAS results for ages 4 and 8 months revealed 58 common SNPs for significant genotypes. 11 common SNPs were identified for body weight, 4 SNPs-for group of traits withers height, rump height, body length, 2 SNPs-for withers height and rump height, 1 SNP-for body length and chest depth. Structural annotation of genomic regions covering a window of ±0.20 Mb showed the presence of 288 genes; 52 of them had the described functions in accordance with gene ontology. The main molecular functions of proteins encoded by these genes are the regulation of transcription, cell proliferation, angiogenesis, body growth, fatty acid and lipid metabolism, nervous system development, and spermatogenesis. SNPs common to body weight and localized within a window of ±200 kb from the structural genes CRADD, HMGA2, MSRB3, FUT8, MAX, and RAB15 were selected to create a test system. The study of meat productivity after slaughter and chemical analysis of muscle tissue in Karachai goats at the age of 8 months of different genotypes according to the identified SNPs revealed that rs268269710 is the most promising for further research and use in breeding. The GG genotype is associated with a larger live weight of animals, a larger carcass yield, the content of the boneless part in it, and the ratio of protein and adipose tissue in meat preferred for dietary nutrition. These results will contribute to the genetic improvement of Karachai goats.

Manual annotation of Drosophila genes: a Genomics Education Partnership protocol.

Annotating the genomes of multiple species allows us to analyze the evolution of their genes. While many eukaryotic genome assemblies already include computational gene predictions, these predictions can benefit from review and refinement through manual gene annotation. The Genomics Education Partnership (GEP; https://thegep.org/) developed a structural annotation protocol for protein-coding genes that enables undergraduate student and faculty researchers to create high-quality gene annotations that can be utilized in subsequent scientific investigations. For example, this protocol has been utilized by the GEP faculty to engage undergraduate students in the comparative annotation of genes involved in the insulin signaling pathway in 27 Drosophila species, using D. melanogaster as the reference genome. Students construct gene models using multiple lines of computational and empirical evidence including expression data (e.g., RNA-Seq), sequence similarity (e.g., BLAST and multiple sequence alignment), and computational gene predictions. Quality control measures require each gene be annotated by at least two students working independently, followed by reconciliation of the submitted gene models by a more experienced student. This article provides an overview of the annotation protocol and describes how discrepancies in student submitted gene models are resolved to produce a final, high-quality gene set suitable for subsequent analyses. The protocol can be adapted to other scientific questions (e.g., expansion of the Drosophila Muller F element) and species (e.g., parasitoid wasps) to provide additional opportunities for undergraduate students to participate in genomics research. These student annotation efforts can substantially improve the quality of gene annotations in publicly available genomic databases.

First draft of an annotated genome for a lichenised strain of the green alga Diplosphaera chodatii (Prasiolales, Trebouxiophyceae)

ABSTRACT Although genome sequences of lichenized fungi are increasingly becoming available, genome sequences of microalgae involved in the lichen symbiosis are still scarce. For lichenized eukaryotic algae, genome sequencing has focused mostly on Trebouxia and Asterochloris, with little genomic data available for Stichococcus-like algae, such as Diplosphaera. The genus Diplosphaera is a common component of biological soil crusts, and often occurs associated with lichens of the family Verrucariaceae. It is characterized by cylindrical to spherical cells containing a plate-like chloroplast, and more specifically by a vegetative cell division that leads to the formation of typical two- to four-celled clusters. Here, we present a draft genome sequence for the algal partner of an Australian lichen specimen of Endocarpon pusillum. The genome was sequenced with Pac Bio long read and Illumina short read technologies, and transcriptome data were generated to inform the structural annotations. This algal strain is here identified as Diplosphaera chodatii based on nuSSU and ITS data. Compared with closely related lichenized and non-lichenized algae, the genome of D. chodatii stands out for its large size (85.6 Mb) and gene content (21,261 protein-encoding regions), as well as its high rate of duplicated genes (60% of the BUSCO genes are duplicated). These results suggest that whole genome duplication or large-scale segmental duplications may have occurred in the evolutionary history of this algal species. HIGHLIGHTS Little genome data are available for lichenized algae. We generated the first genome for a lichenized Diplosphaera chodatii. Results suggest a possible whole genome duplication in this species.

Activity labelled molecular networking fuels the antioxidation active molecules profile of Ginger.

Ginger has been used as consumed food spice and folk medicine in daily life for thousands of years in various regions of the world. Considerable antioxidation is one of the major activities for Ginger to exhibit health-promoting effects. In this study, a bioinformatic workflow was developed to generate activity labelled molecular networking (ALMN) to fuel the antioxidation active molecules profile of Ginger. In ALMN, antioxidation activity data, which was defined as correlation (r and p value) between the relative abundance of a molecule in fractions and the activity level of each fraction, was labelled to feature-based molecular network to profile out antioxidation active molecules visually. Fragmentation tree was further computed as a complementary way to conduct high confidence structure annotations of antioxidation active molecules. Consequently, 48 molecules were prioritized as antioxidation active molecules from 11,720 metabolite molecules of Ginger in a systematical way.

Superoxide dismutase gene family in cassava revealed their involvement in environmental stress via genome-wide analysis

Superoxide dismutase (SOD) is a crucial metal-containing enzyme that plays a vital role in catalyzing the dismutation of superoxide anions, converting them into molecular oxygen and hydrogen peroxide, essential for enhancing plant stress tolerance. We identified 8 SOD genes (4 CSODs, 2 FSODs, and 2 MSODs) in cassava. Bioinformatics analyses provided insights into chromosomal location, phylogenetic relationships, gene structure, conserved motifs, and gene ontology annotations. MeSOD genes were classified into two groups through phylogenetic analysis, revealing evolutionary connections. Promoters of these genes harbored stress-related cis-elements. Duplication analysis indicated the functional significance of MeCSOD2/MeCSOD4 and MeMSOD1/MeMSOD2. Through qRT-PCR, MeCSOD2 responded to salt stress, MeMSOD2 to drought, and cassava bacterial blight. Silencing MeMSOD2 increased XpmCHN11 virulence, indicating MeMSOD2 is essential for cassava's defense against XpmCHN11 infection. These findings enhance our understanding of the SOD gene family's role in cassava and contribute to strategies for stress tolerance improvement.

Uncovering new families and folds in the natural protein universe

We are now entering a new era in protein sequence and structure annotation, with hundreds of millions of predicted protein structures made available through the AlphaFold database1. These models cover nearly all proteins that are known, including those challenging to annotate for function or putative biological role using standard homology-based approaches. In this study, we examine the extent to which the AlphaFold database has structurally illuminated this ‘dark matter’ of the natural protein universe at high predicted accuracy. We further describe the protein diversity that these models cover as an annotated interactive sequence similarity network, accessible at https://uniprot3d.org/atlas/AFDB90v4. By searching for novelties from sequence, structure and semantic perspectives, we uncovered the β-flower fold, added several protein families to Pfam database2 and experimentally demonstrated that one of these belongs to a new superfamily of translation-targeting toxin–antitoxin systems, TumE–TumA. This work underscores the value of large-scale efforts in identifying, annotating and prioritizing new protein families. By leveraging the recent deep learning revolution in protein bioinformatics, we can now shed light into uncharted areas of the protein universe at an unprecedented scale, paving the way to innovations in life sciences and biotechnology.

Multiple structural and functional annotations based in-silico characterization of Q9BRX8 protein

Multiple structural and functional annotations based in-silico characterization of Q9BRX8 protein

Functional and structural annotation of an uncharacterized protein A0A077Z4I5 from Trichuris trichiura

Functional and structural annotation of an uncharacterized protein A0A077Z4I5 from Trichuris trichiura

DeepSAT: Learning Molecular Structures from Nuclear Magnetic Resonance Data

The identification of molecular structure is essential for understanding chemical diversity and for developing drug leads from small molecules. Nevertheless, the structure elucidation of small molecules by Nuclear Magnetic Resonance (NMR) experiments is often a long and non-trivial process that relies on years of training. To achieve this process efficiently, several spectral databases have been established to retrieve reference NMR spectra. However, the number of reference NMR spectra available is limited and has mostly facilitated annotation of commercially available derivatives. Here, we introduce DeepSAT, a neural network-based structure annotation and scaffold prediction system that directly extracts the chemical features associated with molecular structures from their NMR spectra. Using only the 1H-13C HSQC spectrum, DeepSAT identifies related known compounds and thus efficiently assists in the identification of molecular structures. DeepSAT is expected to accelerate chemical and biomedical research by accelerating the identification of molecular structures.

Manual annotation of Drosophila genes: a Genomics Education Partnership protocol

Annotating the genomes of multiple species allows us to analyze the evolution of their genes. While many eukaryotic genome assemblies already include computational gene predictions, these predictions can benefit from review and refinement through manual gene annotation. The Genomics Education Partnership (GEP; https://thegep.org/) developed a structural annotation protocol for protein-coding genes that enables undergraduate student and faculty researchers to create high-quality gene annotations that can be utilized in subsequent scientific investigations. For example, this protocol has been utilized by the GEP faculty to engage undergraduate students in the comparative annotation of genes involved in the insulin signaling pathway in 27 Drosophila species, using D. melanogaster as the reference genome. Students construct gene models using multiple lines of computational and empirical evidence including expression data (e.g., RNA-Seq), sequence similarity (e.g., BLAST and multiple sequence alignment), and computational gene predictions. Quality control measures require each gene be annotated by at least two students working independently, followed by reconciliation of the submitted gene models by a more experienced student. This article provides an overview of the annotation protocol and describes how discrepancies in student submitted gene models are resolved to produce a final, high-quality gene set suitable for subsequent analyses. The protocol can be adapted to other scientific questions (e.g., expansion of the Drosophila Muller F element) and species (e.g., parasitoid wasps) to provide additional opportunities for undergraduate students to participate in genomics research. These student annotation efforts can substantially improve the quality of gene annotations in publicly available genomic databases.

Identification of potential genetic causal variants for obesity-related traits using statistical fine mapping.

Obesity is highly influenced by heritability and variant effects. While previous genome-wide association studies (GWASs) have successfully identified numerous genetic loci associated with obesity-related traits [body mass index (BMI) and waist-to-hip ratio (WHR)], most causal variants remain unidentified. The high degree of linkage disequilibrium (LD) throughout the genome makes it extremely difficult to distinguish the GWAS-associated SNPs that exert a true biological effect. This study was to identify the potential causal variants having a biological effect on obesity-related traits. We used Probabilistic Annotation INTegratOR, a Bayesian fine-mapping method, which incorporated genetic association data (GWAS summary statistics), LD structure, and functional annotations to calculate a posterior probability of causality for SNPs across all loci of interest. Moreover, we performed gene expression analysis using the available public transcriptomic data to validate the corresponding genes of the potential causal SNPs partially. We identified 96 SNPs for BMI and 43 SNPs for WHR with a high posterior probability of causality (> 99%), including 49 BMI SNPs and 24 WHR SNPs which did not reach genome-wide significance in the original GWAS. Finally, we partially validated some genes corresponding to the potential causal SNPs. Using a statistical fine-mapping approach, we identified a set of potential causal variants to be prioritized for future functional validation and also detected some novel trait-associated variants. These results provided novel insight into our understanding of the genetics of obesity and also demonstrated that fine mapping may improve upon the results identified by the original GWASs.

A novel missense compound heterozygous variant in TLR1 gene is associated with susceptibility to rheumatoid arthritis - structural perspective and functional annotations.

Besides human leukocyte antigen (HLA-DRB1) locus, more than 100 loci across the genome have been identified and linked with the onset, expression and/or progression of rheumatoid arthritis (RA). However, there are still grey areas in our understanding of the key genetic contributors of the disease, particularly in familial cases. In the present study, we have performed the whole exome sequencing (WES) of RA patients from two consanguineous families of Pakistan in a quest to identify novel, high-impact, RA-susceptibility genetic variants. Through stepwise filtering, around 17,000 variants (common in the affected members) were recognized, out of which 2651 were predicted to be deleterious. Of these, 196 had direct relevance to RA. When selected for homozygous recessive mode of inheritance, two novel pathogenic variants (c.1324T>C, p.Cys442→Arg442; c.2036T>C, p.Ile679→Thr679) in the TLR1 gene displayed the role of compound heterozygosity in modulating the phenotypic expression and penetrance of RA. The structural and functional consequences of the TLR1 missense single nucleotide mutations (Cys442→Arg442; Ile679→Thr679) were evaluated through molecular dynamic simulation (MDS) studies. Analysis showed domain's rigidification, conferring stability to mutant TLR1-TIR/TIRAP-TIR complex with concomitant increase in molecular interactions with pro-inflammatory cytokines, compared to the wild-type conformation. Gene co-expression network analysis highlighted interlinked partnering genes along with interleukin-6 production of TLR1 (corrected p-value 2.98e-4) and acetylcholine receptor activity of CHRNG (corrected p-value 6.12e-2) as highly enriched associated functions. The results, validated through case-control study subjects, suggested that the variants identified through WES and confirmed through Sanger sequencing and MDS are the novel disease variants and are likely to confer RA-susceptibility, independently and/or in a family-specific context. Key Points • Exploration of population based/ethno-specific big data is imperative to identify novel causal variants of RA. • Two new deleterious missense mutations in mutational hotspot exon 4 of TLR1 gene have been identified in Pakistani RA patients. • MD simulation data provides evidence for domain's rigidification, conferring stability to mutant TLR1-TIR/TIRAP-TIR complex, with concomitant increase in production of pro-inflammatory cytokines, thus adding to the onset/erosive outcome of RA.

Comparative evolutionary analyses of eight whitefly Bemisia tabaci sensu lato genomes: cryptic species, agricultural pests and plant-virus vectors

BackgroundThe group of > 40 cryptic whitefly species called Bemisia tabaci sensu lato are amongst the world’s worst agricultural pests and plant-virus vectors. Outbreaks of B. tabaci s.l. and the associated plant-virus diseases continue to contribute to global food insecurity and social instability, particularly in sub-Saharan Africa and Asia. Published B. tabaci s.l. genomes have limited use for studying African cassava B. tabaci SSA1 species, due to the high genetic divergences between them. Genomic annotations presented here were performed using the ‘Ensembl gene annotation system’, to ensure that comparative analyses and conclusions reflect biological differences, as opposed to arising from different methodologies underpinning transcript model identification.ResultsWe present here six new B. tabaci s.l. genomes from Africa and Asia, and two re-annotated previously published genomes, to provide evolutionary insights into these globally distributed pests. Genome sizes ranged between 616—658 Mb and exhibited some of the highest coverage of transposable elements reported within Arthropoda. Many fewer total protein coding genes (PCG) were recovered compared to the previously published B. tabaci s.l. genomes and structural annotations generated via the uniform methodology strongly supported a repertoire of between 12.8—13.2 × 103 PCG. An integrative systematics approach incorporating phylogenomic analysis of nuclear and mitochondrial markers supported a monophyletic Aleyrodidae and the basal positioning of B. tabaci Uganda-1 to the sub-Saharan group of species. Reciprocal cross-mating data and the co-cladogenesis pattern of the primary obligate endosymbiont ‘Candidatus Portiera aleyrodidarum’ from 11 Bemisia genomes further supported the phylogenetic reconstruction to show that African cassava B. tabaci populations consist of just three biological species. We include comparative analyses of gene families related to detoxification, sugar metabolism, vector competency and evaluate the presence and function of horizontally transferred genes, essential for understanding the evolution and unique biology of constituent B. tabaci. s.l species.ConclusionsThese genomic resources have provided new and critical insights into the genetics underlying B. tabaci s.l. biology. They also provide a rich foundation for post-genomic research, including the selection of candidate gene-targets for innovative whitefly and virus-control strategies.

The repetitive structure of DNA clamps: An overlooked protein tandem repeat

Structured tandem repeats proteins (STRPs) are a specific kind of tandem repeat proteins characterized by a modular and repetitive three-dimensional structure arrangement. The majority of STRPs adopt solenoid structures, but with the increasing availability of experimental structures and high-quality predicted structural models, more STRP folds can be characterized. Here, we describe “Box repeats”, an overlooked STRP fold present in the DNA sliding clamp processivity factors, which has eluded classification although structural data has been available since the late 1990s. Each Box repeat is a β⍺βββ module of about 60 residues, which forms a class V “beads-on-a-string” type STRP. The number of repeats present in processivity factors is organism dependent. Monomers of PCNA proteins in both Archaea and Eukarya have 4 repeats, while the monomers of bacterial beta-sliding clamps have 6 repeats. This new repeat fold has been added to the RepeatsDB database, which now provides structural annotation for 66 Box repeat proteins belonging to different organisms, including viruses.

Spatially resolved multiomics of human cardiac niches

The function of a cell is defined by its intrinsic characteristics and its niche: the tissue microenvironment in which it dwells. Here we combine single-cell and spatial transcriptomics data to discover cellular niches within eight regions of the human heart. We map cells to microanatomical locations and integrate knowledge-based and unsupervised structural annotations. We also profile the cells of the human cardiac conduction system1. The results revealed their distinctive repertoire of ion channels, G-protein-coupled receptors (GPCRs) and regulatory networks, and implicated FOXP2 in the pacemaker phenotype. We show that the sinoatrial node is compartmentalized, with a core of pacemaker cells, fibroblasts and glial cells supporting glutamatergic signalling. Using a custom CellPhoneDB.org module, we identify trans-synaptic pacemaker cell interactions with glia. We introduce a druggable target prediction tool, drug2cell, which leverages single-cell profiles and drug–target interactions to provide mechanistic insights into the chronotropic effects of drugs, including GLP-1 analogues. In the epicardium, we show enrichment of both IgG+ and IgA+ plasma cells forming immune niches that may contribute to infection defence. Overall, we provide new clarity to cardiac electro-anatomy and immunology, and our suite of computational approaches can be applied to other tissues and organs.

Analysis of the Genomic Features and Evolutionary History of Pithovirus-Like Isolates Reveals Two Major Divergent Groups of Viruses.

New representatives of the phylum Nucleocytoviricota have been rapidly described in the last decade. Despite this, not all viruses of this phylum are allocated to recognized taxonomic families, as is the case for orpheovirus, pithovirus, and cedratvirus, which form the proposed family Pithoviridae. In this study, we performed comprehensive comparative genomic analyses of 8 pithovirus-like isolates, aiming to understand their common traits and evolutionary history. Structural and functional genome annotation was performed de novo for all the viruses, which served as a reference for pangenome construction. The synteny analysis showed substantial differences in genome organization between these viruses, with very few and short syntenic blocks shared between orpheovirus and its relatives. It was possible to observe an open pangenome with a significant increase in the slope when orpheovirus was added, alongside a decrease in the core genome. Network analysis placed orpheovirus as a distant and major hub with a large fraction of unique clusters of orthologs, indicating a distant relationship between this virus and its relatives, with only a few shared genes. Additionally, phylogenetic analyses of strict core genes shared with other viruses of the phylum reinforced the divergence of orpheovirus from pithoviruses and cedratviruses. Altogether, our results indicate that although pithovirus-like isolates share common features, this group of ovoid-shaped giant viruses presents substantial differences in gene contents, genomic architectures, and the phylogenetic history of several core genes. Our data indicate that orpheovirus is an evolutionarily divergent viral entity, suggesting its allocation to a different viral family, Orpheoviridae. IMPORTANCE Giant viruses that infect amoebae form a monophyletic group named the phylum Nucleocytoviricota. Despite being genomically and morphologically very diverse, the taxonomic categories of some clades that form this phylum are not yet well established. With advances in isolation techniques, the speed at which new giant viruses are described has increased, escalating the need to establish criteria to define the emerging viral taxa. In this work, we performed a comparative genomic analysis of representatives of the putative family Pithoviridae. Based on the dissimilarity of orpheovirus from the other viruses of this putative family, we propose that orpheovirus be considered a member of an independent family, Orpheoviridae, and suggest criteria to demarcate families consisting of ovoid-shaped giant viruses.

Artificial Intelligence for context-aware surgical guidance in complex robot-assisted oncological procedures: An exploratory feasibility study

IntroductionComplex oncological procedures pose various surgical challenges including dissection in distinct tissue planes and preservation of vulnerable anatomical structures throughout different surgical phases. In rectal surgery, violation of dissection planes increases the risk of local recurrence and autonomous nerve damage resulting in incontinence and sexual dysfunction. This work explores the feasibility of phase recognition and target structure segmentation in robot-assisted rectal resection (RARR) using machine learning. Materials and methodsA total of 57 RARR were recorded and subsets of these were annotated with respect to surgical phases and exact locations of target structures (anatomical structures, tissue types, static structures, and dissection areas). For surgical phase recognition, three machine learning models were trained: LSTM, MSTCN, and Trans-SVNet. Based on pixel-wise annotations of target structures in 9037 images, individual segmentation models based on DeepLabv3 were trained. Model performance was evaluated using F1 score, Intersection-over-Union (IoU), accuracy, precision, recall, and specificity. ResultsThe best results for phase recognition were achieved with the MSTCN model (F1 score: 0.82 ± 0.01, accuracy: 0.84 ± 0.03). Mean IoUs for target structure segmentation ranged from 0.14 ± 0.22 to 0.80 ± 0.14 for organs and tissue types and from 0.11 ± 0.11 to 0.44 ± 0.30 for dissection areas. Image quality, distorting factors (i.e. blood, smoke), and technical challenges (i.e. lack of depth perception) considerably impacted segmentation performance. ConclusionMachine learning-based phase recognition and segmentation of selected target structures are feasible in RARR. In the future, such functionalities could be integrated into a context-aware surgical guidance system for rectal surgery.