Biology subjects

Savas, B.

Publications and source records attributed to Savas, B..

6 recordsLinked to original sources

bioRxiv2025

Exploring the Potential of AlphaFold Distograms for Flexibility Assignment in Cryo-EM Maps

AlphaFold models provide static structural predictions, limiting their use in interpreting flexible regions in low-resolution cryo-EM maps. Here, we assess whether AlphaFold-generated distograms can instead reveal conformational flexibility, focusing on binding-induced hinge motions. For this, we examined the key metabolic AK2/AIFM1 complex, where molecular dynamics and cryo-EM confirm a hinge motion in AK2 upon binding. Notably, this motion is captured in the AlphaFold2/3 distograms of apo AK2, even though it is absent in the predicted structures. By extending our analysis to other systems, we demonstrate that distograms may offer a valuable, model-independent method for interpreting ambiguous regions in cryo-EM maps.

bioinformatics↗

bioRxiv2025

Evaluation of the Prognostic Value of Four ECM-Associated Genes in Recurrence and Metastasis of Colorectal Cancer

Colorectal cancer is a significant public health issue due to its high incidence and high mortality rate, especially when diagnosed late. Identifying new biomarkers that can predict the diagnosis and prognosis of CRC is critical to reducing the mortality rate of the disease. This study tested the relationship of four extracellular matrix genes (COL1A1, COL5A1, THBS2 and FN1) with CRC prognosis at the mRNA and protein levels. Our previous studies identified these genes as candidate prognostic biomarkers for CRC. They were analysed in two cohorts: retrospective (130 CRC tumours and 70 healthy colon tissues) and prospective (160 paired tumours and normal tissues with 180 CRC serum samples). Accordingly, in the retrospective cohort, COL1A1 (p>0.0001), COL5A1 (p>0.0001), THBS2 (0.0001) and FN1 (0.001) were expressed at higher levels in tumours compared to controls by qRT-PCR and in metastatic cases compared to non-metastatic cases. In the same cohort, immunohistochemistry analysis showed that increasing protein levels of COL1A1 and FN1 correlated with mRNA levels. In addition, high COL1A1 mRNA levels were associated with OS (p>0.011), RFS (p>0.0001), and DMFS (p>0.0001), high COL5A1 mRNA levels were associated with OS (0.042), RFS (0.013), DMFS (p>0.003) and high FN1 levels were associated with DMFS (p>0.039). In the prospective cohort, COL1A1, COL5A1 and THBS2 were found to be higher in tumour tissue compared to matched normal tissue and in metastatic tumours compared to primary tumours by qRT-PCR in 160 tumours and matched normal tissues. In this cohort, FN1 levels were unexpectedly higher in normal tissues than in tumour tissues. In the prospective cohort, where 180 CRC serum samples were also tested by ELISA, COL1A1 (p>0.0001), COL5A1 (p>0.05), and THBS2 (p>0.05) levels were higher in metastatic cases than in the sera of non-metastatic cases. As a result, this study associated COL1A1, COL5A1 and THBS2 genes, which have different data on their prognostic effects in the literature, with poor prognosis of CRC at both mRNA and protein levels.

cancer biology↗

bioRxiv2024

Towards a greener AlphaFold2 protocol for protein complex modeling: Insights from CAPRI Round 55

In the 55th round of CAPRI, we used enhanced AlphaFold2 (AF2) sampling and data-driven docking. Our AF2 protocol relies on Wallners massive sampling approach, which combines different AF2 versions and sampling parameters to produce thousands of models per target. For T231 (an antibody peptide complex) and T232 (PP2A:TIPRL complex), we employed a 50-fold reduced MinnieFold sampling and a custom ranking approach, leading to a top-ranking medium prediction in both cases. For T233 and T234 (two antibody bound MHC I complexes), we followed data-driven docking, which did not lead to an acceptable model. Our post-CAPRI55 analysis showed that if we would have used our MinnieFold approach on T233 and T234, we could have submitted a medium-quality model for T233 as well. In the scoring challenge, we utilized the scoring function of FoldX, which was effective in selecting acceptable models for T231 and medium quality models for T232. Our success, especially in predicting and ranking a medium quality model for T231 and potentially for T233 underscores the feasibility of green and accurate enhanced AF2 sampling in antibody complex prediction.

bioinformatics↗

bioRxiv2023

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 as well as senescence and death. However, the mechanism by which YPEL2, or YPEL proteins, mediates its effects is yet 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. We here 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 largely 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 RNA binding protein ELAVL1 and selective autophagy receptor SQSTM1. We also found that YPEL2 participates in events associated with the formation/disassembly of stress granules in response to sodium arsenite an oxidative stress inducer. 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.

cell biology↗

bioRxiv2022

Truncated KDM6A exhibits differential chromatin occupancy and regulatory interactions

Epigenetic deregulation is a critical theme which needs further investigation for bladder cancer research. One of the highly mutated genes in bladder cancer is KDM6A, functioning as a H3K27 demethylase and is part of the MLL3/4 complexes. To decipher the role of KDM6A in normal versus tumor setting, we identified the genomic landscape of KDM6A in normal, immortalized and cancer bladder cells. Our results showed differential KDM6A occupancy at the genes involved in cell differentiation, chromatin organization and Notch signaling depending on the cell type and the mutation status of KDM6A. Transcription factor motif analysis revealed HES1 to be enriched at KDM6A peaks identified for T24 bladder cancer cell line, which has a truncating mutation in KDM6A, lacking demethylase domain. Our co-immunoprecipitation experiments reveal TLE co-repressors and HES1 as potential truncated and wild type KDM6A interactors. With the aid of structural modeling, we explored how the truncated KDM6A could interact with TLE, HES1, as well RUNX, HHEX transcription factors. These structures provide a solid mean to study the functions of KDM6A independent of its demethylase activity. Collectively, our work provides important contributions to the understanding of KDM6A malfunction in bladder cancer.

cancer biology↗

bioRxiv2022

Benchmarking the Widely Used Structure-based Binding Affinity Predictors on the 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). Also, when we simplified the prediction problem to a binary classification, i.e., whether a mutation is enriching or depleting the binding, we showed that the highest accuracy is achieved by FoldX with 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 also 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/

bioinformatics↗