Transposase-DNA Complex Structures Reveal Mechanisms for Conjugative Transposition of Antibiotic Resistance.

Conjugative transposition drives the emergence of multidrug resistance in diverse bacterial pathogens, yet the mechanisms are poorly characterized. The Tn1549 conjugative transposon propagates resistance to the antibiotic vancomycin used for severe drug-resistant infections. Here, we present four high-resolution structures of the conserved Y-transposase of Tn1549 complexed with circular transposon DNA intermediates. The structures reveal individual transposition steps and explain how specific DNA distortion and cleavage mechanisms enable DNA strand exchange with an absolute minimum homology requirement. This appears to uniquely allow Tn916-like conjugative transposons to bypass DNA homology and insert into diverse genomic sites, expanding gene transfer. We further uncover a structural regulatory mechanism that prevents premature cleavage of the transposon DNA before a suitable target DNA is found and generate a peptide antagonist that interferes with the transposase-DNA structure to block transposition. Our results reveal mechanistic principles of conjugative transposition that could help control the spread of antibiotic resistance genes.

Benchmarking the accuracy of structure-based binding affinity predictors on Spike-ACE2 deep mutational interaction set.

Since the start of COVID-19 pandemic, a huge effort has been devoted to understanding the Spike (SARS-CoV-2)-ACE2 recognition mechanism. To this end, two deep mutational scanning studies traced the impact of all possible mutations across receptor binding domain (RBD) of Spike and catalytic domain of human ACE2. By concentrating on the interface mutations of these experimental data, we benchmarked six commonly used structure-based binding affinity predictors (FoldX, EvoEF1, MutaBind2, SSIPe, HADDOCK, and UEP). These predictors were selected based on their user-friendliness, accessibility, and speed. As a result of our benchmarking efforts, we observed that none of the methods could generate a meaningful correlation with the experimental binding data. The best correlation is achieved by FoldX (R = -0.51). When we simplified the prediction problem to a binary classification, that is, whether a mutation is enriching or depleting the binding, we showed that the highest accuracy is achieved by FoldX with a 64% success rate. Surprisingly, on this set, simple energetic scoring functions performed significantly better than the ones using extra evolutionary-based terms, as in Mutabind and SSIPe. Furthermore, we demonstrated that recent AI approaches, mmCSM-PPI and TopNetTree, yielded comparable performances to the force field-based techniques. These observations suggest plenty of room to improve the binding affinity predictors in guessing the variant-induced binding profile changes of a host-pathogen system, such as Spike-ACE2. To aid such improvements we provide our benchmarking data at https://github.com/CSB-KaracaLab/RBD-ACE2-MutBench with the option to visualize our mutant models at https://rbd-ace2-mutbench.github.io/.

The impact of AI-based modeling on the accuracy of protein assembly prediction: Insights from CASP15.

In CASP15, 87 predictors submitted around 11 000 models on 41 assembly targets. The community demonstrated exceptional performance in overall fold and interface contact predictions, achieving an impressive success rate of 90% (compared to 31% in CASP14). This remarkable accomplishment is largely due to the incorporation of DeepMind's AF2-Multimer approach into custom-built prediction pipelines. To evaluate the added value of participating methods, we compared the community models to the baseline AF2-Multimer predictor. In over 1/3 of cases, the community models were superior to the baseline predictor. The main reasons for this improved performance were the use of custom-built multiple sequence alignments, optimized AF2-Multimer sampling, and the manual assembly of AF2-Multimer-built subcomplexes. The best three groups, in order, are Zheng, Venclovas, and Wallner. Zheng and Venclovas reached a 73.2% success rate over all (41) cases, while Wallner attained 69.4% success rate over 36 cases. Nonetheless, challenges remain in predicting structures with weak evolutionary signals, such as nanobody-antigen, antibody-antigen, and viral complexes. Expectedly, modeling large complexes also remains challenging due to their high memory compute demands. In addition to the assembly category, we assessed the accuracy of modeling interdomain interfaces in the tertiary structure prediction targets. Models on seven targets featuring 17 unique interfaces were analyzed. Best predictors achieved a 76.5% success rate, with the UM-TBM group being the leader. In the interdomain category, we observed that the predictors faced challenges, as in the case of the assembly category, when the evolutionary signal for a given domain pair was weak or the structure was large. Overall, CASP15 witnessed unprecedented improvement in interface modeling, reflecting the AI revolution seen in CASP14.

Breaking the conformational ensemble barrier: Ensemble structure modeling challenges in CASP15.

For the first time, the 2022 CASP (Critical Assessment of Structure Prediction) community experiment included a section on computing multiple conformations for protein and RNA structures. There was full or partial success in reproducing the ensembles for four of the nine targets, an encouraging result. For protein structures, enhanced sampling with variations of the AlphaFold2 deep learning method was by far the most effective approach. One substantial conformational change caused by a single mutation across a complex interface was accurately reproduced. In two other assembly modeling cases, methods succeeded in sampling conformations near to the experimental ones even though environmental factors were not included in the calculations. An experimentally derived flexibility ensemble allowed a single accurate RNA structure model to be identified. Difficulties included how to handle sparse or low-resolution experimental data and the current lack of effective methods for modeling RNA/protein complexes. However, these and other obstacles appear addressable.

An alternatively spliced PD-L1 isoform PD-L1∆3, and PD-L2 expression in breast cancers: implications for eligibility scoring and immunotherapy response.

Targeting PD-1/PD-L1 has shown substantial therapeutic response and unprecedented long-term durable responses in the clinic. However, several challenges persist, encompassing the prediction of treatment effectiveness and patient responses, the emergence of treatment resistance, and the necessity for additional biomarkers. Consequently, we comprehensively explored the often-overlooked isoforms of crucial immunotherapy players, leveraging transcriptomic analysis, structural modeling, and immunohistochemistry (IHC) data. Our investigation has led to the identification of an alternatively spliced isoform of PD-L1 that lacks exon 3 (PD-L1∆3) and the IgV domain required to interact with PD-1. PD-L1∆3 is expressed more than the canonical isoform in a subset of breast cancers and other TCGA tumors. Using the deep learning-based protein modeling tool AlphaFold2, we show the lack of a possible interaction between PD-L1∆3 and PD-1. In addition, we present data on the expression of an additional ligand for PD-1, PD-L2. PD-L2 expression is widespread and positively correlates with PD-L1 levels in breast and other tumors. We report enriched epithelial-mesenchymal transition (EMT) signature in high PD-L2 transcript expressing (PD-L2 > PD-L1) tumors in all breast cancer subtypes, highlighting potential crosstalk between EMT and immune evasion. Notably, the estrogen gene signature is downregulated in ER + breast tumors with high PD-L2. The data on PD-L2 IHC positivity but PD-L1 negativity in breast tumors, together with our results on PD-L1∆3, highlight the need to utilize PD-L2 and PD-L1 isoform-specific antibodies for staining patient tissue sections to offer a more precise prediction of the outcomes of PD-1/PD-L1 immunotherapy.

Molecular Mechanism of Nucleosome Recognition by the Pioneer Transcription Factor Sox.

Pioneer transcription factors (PTFs) have the remarkable ability to directly bind to chromatin to stimulate vital cellular processes. In this work, we dissect the universal binding mode of Sox PTF by combining extensive molecular simulations and physiochemistry approaches, along with DNA footprinting techniques. As a result, we show that when Sox consensus DNA is located at the solvent-facing DNA strand, Sox binds to the compact nucleosome without imposing any significant conformational changes. We also reveal that the base-specific Sox:DNA interactions (base reading) and Sox-induced DNA changes (shape reading) are concurrently required for sequence-specific nucleosomal DNA recognition. Among three different nucleosome positions located on the positive DNA arm, a sequence-specific reading mechanism is solely satisfied at the superhelical location 2 (SHL2). While SHL2 acts transparently for solvent-facing Sox binding, among the other two positions, SHL4 permits only shape reading. The final position, SHL0 (dyad), on the other hand, allows no reading mechanism. These findings demonstrate that Sox-based nucleosome recognition is essentially guided by intrinsic nucleosome properties, permitting varying degrees of DNA recognition.

A C-term truncated EIF2Bγ protein encoded by an intronically polyadenylated isoform introduces unfavorable EIF2Bγ-EIF2γ interactions.

Eukaryotic translation initiates upon recruitment of the EIF2-GTP·Met-tRNAi ternary complex (TC) to the ribosomes. EIF2 (α, ß, γ subunits) is a GTPase. The GDP to GTP exchange within EIF2 is facilitated by the guanine nucleotide exchange factor EIF2B (α-ε subunits). During stress-induced conditions, phosphorylation of the α-subunit of EIF2 turns EIF2 into an inhibitor of EIF2B. In turn, inhibition of EIF2B decreases TC formation and triggers the internal stress response (ISR), which determines the cell fate. Deregulated ISR has been linked to neurodegenerative disorders and cancer, positioning EIF2B as a promising therapeutic target. Hence, a better understanding of the mechanisms/factors that regulate EIF2B activity is required. Here, combining transcript and protein level analyses, we describe an intronically polyadenylated (IPA) transcript of EIF2B's γ-subunit. We show that the IPA mRNA isoform is translated into a C-terminus truncated protein. Using structural modeling, we predict that the truncated EIF2Bγ protein has unfavorable interactions with EIF2γ, leading to a potential decrease in the stability of the nonproductive EIF2:EIF2B complex. While we discovered and confirmed the IPA mRNA isoform in breast cancer cells, the expression of this isoform is not cancer-specific and is widely present in normal tissues. Overall, our data show that a truncated EIF2Bγ protein co-exists with the canonical protein and is an additional player to regulate the equilibrium between productive and nonproductive states of the EIF2:EIF2B complex. These results may have implications in stress-induced translation control in normal and disease states. Our combinatorial approach demonstrates the need to study noncanonical mRNA and protein isoforms to understand protein interactions and intricate molecular mechanisms.

Enzymatic Generation of Double-Modified AdoMet Analogues and Their Application in Cascade Reactions with Different Methyltransferases.

Methyltransferases (MTases) have become an important tool for site-specific alkylation and biomolecular labelling. In biocatalytic cascades with methionine adenosyltransferases (MATs), transfer of functional moieties has been realized starting from methionine analogues and ATP. However, the widespread use of S-adenosyl-l-methionine (AdoMet) and the abundance of MTases accepting sulfonium centre modifications limit selective modification in mixtures. AdoMet analogues with additional modifications at the nucleoside moiety bear potential for acceptance by specific MTases. Here, we explored the generation of double-modified AdoMets by an engineered Methanocaldococcus jannaschii MAT (PC-MjMAT), using 19 ATP analogues in combination with two methionine analogues. This substrate screening was extended to cascade reactions and to MTase competition assays. Our results show that MTase targeting selectivity can be improved by using bulky substituents at the N6 of adenine. The facile access to >10 new AdoMet analogues provides the groundwork for developing MAT-MTase cascades for orthogonal biomolecular labelling.

Modeling the Dynamics of Protein-Protein Interfaces, How and Why?

Protein-protein assemblies act as a key component in numerous cellular processes. Their accurate modeling at the atomic level remains a challenge for structural biology. To address this challenge, several docking and a handful of deep learning methodologies focus on modeling protein-protein interfaces. Although the outcome of these methods has been assessed using static reference structures, more and more data point to the fact that the interaction stability and specificity is encoded in the dynamics of these interfaces. Therefore, this dynamics information must be taken into account when modeling and assessing protein interactions at the atomistic scale. Expanding on this, our review initially focuses on the recent computational strategies aiming at investigating protein-protein interfaces in a dynamic fashion using enhanced sampling, multi-scale modeling, and experimental data integration. Then, we discuss how interface dynamics report on the function of protein assemblies in globular complexes, in fuzzy complexes containing intrinsically disordered proteins, as well as in active complexes, where chemical reactions take place across the protein-protein interface.

Assessment of the CASP14 assembly predictions.

In CASP14, 39 research groups submitted more than 2500 3D models on 22 protein complexes. In general, the community performed well in predicting the fold of the assemblies (for 80% of the targets), although it faced significant challenges in reproducing the native contacts. This is especially the case for the complexes without whole-assembly templates. The leading predictor, BAKER-experimental, used a methodology combining classical techniques (template-based modeling, protein docking) with deep learning-based contact predictions and a fold-and-dock approach. The Venclovas team achieved the runner-up position with template-based modeling and docking. By analyzing the target interfaces, we showed that the complexes with depleted charged contacts or dominating hydrophobic interactions were the most challenging ones to predict. We also demonstrated that if AlphaFold2 predictions were at hand, the interface prediction challenge could be alleviated for most of the targets. All in all, it is evident that new approaches are needed for the accurate prediction of assemblies, which undoubtedly will expand on the significant improvements in the tertiary structure prediction field.

M3: an integrative framework for structure determination of molecular machines.

We present a broadly applicable, user-friendly protocol that incorporates sparse and hybrid experimental data to calculate quasi-atomic-resolution structures of molecular machines. The protocol uses the HADDOCK framework, accounts for extensive structural rearrangements both at the domain and atomic levels and accepts input from all structural and biochemical experiments whose data can be translated into interatomic distances and/or molecular shapes.

Novel mutations in KMT2B offer pathophysiological insights into childhood-onset progressive dystonia.

Rapid progress has recently been made in the elucidation of the genetic basis of childhood-onset inherited generalized dystonia (IGD) due to the implementation of genomic sequencing methodologies. We identified four patients with childhood-onset IGD harboring novel disease-causing mutations in lysine-specific histone methyltransferase 2B gene (KMT2B) by whole-exome sequencing. The main focus of this paper is to gain novel pathophysiological insights through understanding the molecular consequences of these mutations. The disease course is mostly progressive, evolving from lower limbs into generalized dystonia, which could be associated with dysarthria, dysphonia, intellectual disability, orofacial dyskinesia, and sometimes distinct dysmorphic facial features. In two patients, motor performances improved after bilateral implantation of deep brain stimulation in the globus pallidus internus (GPi-DBS). Pharmacotherapy with trihexyphenidyl reduced dystonia in two patients. We discovered three novel KMT2B mutations. Our analyses revealed that the mutation in patient 1 (c.7463A > G, p.Y2488C) is localized in the highly conserved FYRC domain of KMT2B. This mutation holds the potential to alter the inter-domain FYR interactions, which could lead to KMT2B instability. The mutations in patients 2 and 3 (c.3596_3697insC, p.M1202Dfs*22; c.4229delA, p.Q1410Rfs*12) lead to predicted unstable transcripts, likely to be subject to degradation by non-sense-mediated decay. Childhood-onset progressive dystonia with orofacial involvement is one of the main clinical manifestations of KMT2B mutations. In all, 26% (18/69) of the reported cases have T2 signal alterations of the globus pallidus internus, mostly at a younger age. Anticholinergic medication and GPi-DBS are promising treatment options and shall be considered early.

Correction to: Novel mutations in KMT2B offer pathophysiological insights on childhood-onset progressive dystonia.

Rapid progress has recently been made in the elucidation of the genetic basis of childhood-onset inherited generalized dystonia (IGD) due to the implementation of genomic sequencing methodologies. We identified four patients with childhood-onset IGD harboring novel disease-causing mutations in lysine-specific histone methyltransferase 2B gene (KMT2B) by whole-exome sequencing. The main focus of this paper is to gain novel pathophysiological insights through understanding the molecular consequences of these mutations.The disease course is mostly progressive, evolving from lower limbs into generalized dystonia, which could be associated with dysarthria, dysphonia, intellectual disability, orofacial dyskinesia, and sometimes distinct dysmorphic facial features. In two patients, motor performances improved after bilateral implantation of deep brain stimulation in the globus pallidus internus (GPi-DBS). Pharmacotherapy with trihexyphenidyl reduced dystonia in two patients.We discovered three novel KMT2B mutations. Our analyses revealed that the mutation in patient 1 (c.7463 A > G, p.Y2488C) is localized in the highly conserved FYRC domain of KMT2B. This mutation holds the potential to alter the inter-domain FYR interactions, which could lead to KMT2B instability. The mutations in patients 2 and 3 (c.3602dupC, p.M1202Dfs*22; c.4229delA, p.Q1410Rfs*12) lead to predicted unstable transcripts, likely to be subject to degradation by non-sense mediated decay.Childhood-onset progressive dystonia with orofacial involvement is one of the main clinical manifestations of KMT2B mutations. In all, 26% (18/69) of the reported cases have T2 signal alterations of the globus pallidus internus, mostly at a younger age. Anticholinergic medication and GPi-DBS are promising treatment options and shall be considered early.An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Prediction of homoprotein and heteroprotein complexes by protein docking and template-based modeling: A CASP-CAPRI experiment.

We present the results for CAPRI Round 30, the first joint CASP-CAPRI experiment, which brought together experts from the protein structure prediction and protein-protein docking communities. The Round comprised 25 targets from amongst those submitted for the CASP11 prediction experiment of 2014. The targets included mostly homodimers, a few homotetramers, and two heterodimers, and comprised protein chains that could readily be modeled using templates from the Protein Data Bank. On average 24 CAPRI groups and 7 CASP groups submitted docking predictions for each target, and 12 CAPRI groups per target participated in the CAPRI scoring experiment. In total more than 9500 models were assessed against the 3D structures of the corresponding target complexes. Results show that the prediction of homodimer assemblies by homology modeling techniques and docking calculations is quite successful for targets featuring large enough subunit interfaces to represent stable associations. Targets with ambiguous or inaccurate oligomeric state assignments, often featuring crystal contact-sized interfaces, represented a confounding factor. For those, a much poorer prediction performance was achieved, while nonetheless often providing helpful clues on the correct oligomeric state of the protein. The prediction performance was very poor for genuine tetrameric targets, where the inaccuracy of the homology-built subunit models and the smaller pair-wise interfaces severely limited the ability to derive the correct assembly mode. Our analysis also shows that docking procedures tend to perform better than standard homology modeling techniques and that highly accurate models of the protein components are not always required to identify their association modes with acceptable accuracy. Proteins 2016; 84(Suppl 1):323-348. © 2016 Wiley Periodicals, Inc.

Blind prediction of interfacial water positions in CAPRI.

We report the first assessment of blind predictions of water positions at protein-protein interfaces, performed as part of the critical assessment of predicted interactions (CAPRI) community-wide experiment. Groups submitting docking predictions for the complex of the DNase domain of colicin E2 and Im2 immunity protein (CAPRI Target 47), were invited to predict the positions of interfacial water molecules using the method of their choice. The predictions-20 groups submitted a total of 195 models-were assessed by measuring the recall fraction of water-mediated protein contacts. Of the 176 high- or medium-quality docking models-a very good docking performance per se-only 44% had a recall fraction above 0.3, and a mere 6% above 0.5. The actual water positions were in general predicted to an accuracy level no better than 1.5 Å, and even in good models about half of the contacts represented false positives. This notwithstanding, three hotspot interface water positions were quite well predicted, and so was one of the water positions that is believed to stabilize the loop that confers specificity in these complexes. Overall the best interface water predictions was achieved by groups that also produced high-quality docking models, indicating that accurate modelling of the protein portion is a determinant factor. The use of established molecular mechanics force fields, coupled to sampling and optimization procedures also seemed to confer an advantage. Insights gained from this analysis should help improve the prediction of protein-water interactions and their role in stabilizing protein complexes.

Advances in integrative modeling of biomolecular complexes.

High-resolution structural information is needed in order to unveil the underlying mechanistic of biomolecular function. Due to the technical limitations or the nature of the underlying complexes, acquiring atomic resolution information is difficult for many challenging systems, while, often, low-resolution biochemical or biophysical data can still be obtained. To make best use of all the available information and shed light on these challenging systems, integrative computational tools are required that can judiciously combine and accurately translate sparse experimental data into structural information. In this review we discuss the current state of integrative approaches, the challenges they are confronting and the advances made regarding those challenges. Recent developments are underpinned by noteworthy application examples taken from the literature. Within this context, we also position our data-driven docking approach, HADDOCK that can integrate a variety of information sources to drive the modeling of biomolecular complexes. Only a synergistic combination of experiment and modeling will allow us to tackle the challenges of adding the structural dimension to interactomes, shed "atomic" light onto molecular processes and understand the underlying mechanistic of biomolecular function. The current state of integrative approaches indicates that they are poised to take those challenges.

Dynamic proximity interaction profiling suggests that YPEL2 is involved in cellular stress surveillance.

YPEL2 is a member of the evolutionarily conserved YPEL family involved in cellular proliferation, mobility, differentiation, senescence, and death. However, the mechanism by which YPEL2, or YPEL proteins, mediates its effects is largely unknown. Proteins perform their functions in a network of proteins whose identities, amounts, and compositions change spatiotemporally in a lineage-specific manner in response to internal and external stimuli. Here, we explored interaction partners of YPEL2 by using dynamic TurboID-coupled mass spectrometry analyses to infer a function for the protein. Our results using inducible transgene expressions in COS7 cells indicate that proximity interaction partners of YPEL2 are mainly involved in RNA and mRNA metabolic processes, ribonucleoprotein complex biogenesis, regulation of gene silencing by miRNA, and cellular responses to stress. We showed that YPEL2 interacts with the RNA-binding protein ELAVL1 and the selective autophagy receptor SQSTM1. We also found that YPEL2 localizes stress granules in response to sodium arsenite, an oxidative stress inducer, which suggests that YPEL2 participates in stress granule-related processes. Establishing a point of departure in the delineation of structural/functional features of YPEL2, our results suggest that YPEL2 may be involved in stress surveillance mechanisms.

The HADDOCK2.4 web server for integrative modeling of biomolecular complexes.

Interactions between macromolecules, such as proteins and nucleic acids, are essential for cellular functions. Experimental methods can fail to provide all the information required to fully model biomolecular complexes at atomic resolution, particularly for large and heterogeneous assemblies. Integrative computational approaches have, therefore, gained popularity, complementing traditional experimental methods in structural biology. Here, we introduce HADDOCK2.4, an integrative modeling platform, and its updated web interface ( https://wenmr.science.uu.nl/haddock2.4 ). The platform seamlessly integrates diverse experimental and theoretical data to generate high-quality models of macromolecular complexes. The user-friendly web server offers automated parameter settings, access to distributed computing resources, and pre- and post-processing steps that enhance the user experience. To present the web server's various interfaces and features, we demonstrate two different applications: (i) we predict the structure of an antibody-antigen complex by using NMR data for the antigen and knowledge of the hypervariable loops for the antibody, and (ii) we perform coarse-grained modeling of PRC1 with a nucleosome particle guided by mutagenesis and functional data. The described protocols require some basic familiarity with molecular modeling and the Linux command shell. This new version of our widely used HADDOCK web server allows structural biologists and non-experts to explore intricate macromolecular assemblies encompassing various molecule types.

On the usefulness of ion-mobility mass spectrometry and SAXS data in scoring docking decoys.

Scoring, the process of selecting the biologically relevant solution from a pool of generated conformations, is one of the major challenges in the field of biomolecular docking. A prominent way to cope with this challenge is to incorporate information-based terms into the scoring function. Within this context, low-resolution shape data obtained from either ion-mobility mass spectrometry (IM-MS) or SAXS experiments have been integrated into the conventional scoring function of the information-driven docking program HADDOCK. Here, the strengths and weaknesses of IM-MS-based and SAXS-based scoring, either in isolation or in combination with the HADDOCK score, are systematically assessed. The results of an analysis of a large docking decoy set composed of dimers generated by running HADDOCK in ab initio mode reveal that the content of the IM-MS data is of too low resolution for selecting correct models, while scoring with SAXS data leads to a significant improvement in performance. However, the effectiveness of SAXS scoring depends on the shape and the arrangement of the complex, with prolate and oblate systems showing the best performance. It is observed that the highest accuracy is achieved when SAXS scoring is combined with the energy-based HADDOCK score.

The Impact of AI-Based Modeling on the Accuracy of Protein Assembly Prediction: Insights from CASP15.

In CASP15, 87 predictors submitted around 11,000 models on 41 assembly targets. The community demonstrated exceptional performance in overall fold and interface contact prediction, achieving an impressive success rate of 90% (compared to 31% in CASP14). This remarkable accomplishment is largely due to the incorporation of DeepMind's AF2-Multimer approach into custom-built prediction pipelines. To evaluate the added value of participating methods, we compared the community models to the baseline AF2-Multimer predictor. In over 1/3 of cases the community models were superior to the baseline predictor. The main reasons for this improved performance were the use of custom-built multiple sequence alignments, optimized AF2-Multimer sampling, and the manual assembly of AF2-Multimer-built subcomplexes. The best three groups, in order, are Zheng, Venclovas and Wallner. Zheng and Venclovas reached a 73.2% success rate over all (41) cases, while Wallner attained 69.4% success rate over 36 cases. Nonetheless, challenges remain in predicting structures with weak evolutionary signals, such as nanobody-antigen, antibody-antigen, and viral complexes. Expectedly, modeling large complexes remains also challenging due to their high memory compute demands. In addition to the assembly category, we assessed the accuracy of modeling interdomain interfaces in the tertiary structure prediction targets. Models on seven targets featuring 17 unique interfaces were analyzed. Best predictors achieved the 76.5% success rate, with the UM-TBM group being the leader. In the interdomain category, we observed that the predictors faced challenges, as in the case of the assembly category, when the evolutionary signal for a given domain pair was weak or the structure was large. Overall, CASP15 witnessed unprecedented improvement in interface modeling, reflecting the AI revolution seen in CASP14.