Dr. Melike Dönertaş will give a seminar on “Temporal changes in gene expression across individuals, organs, and cells” on 17 January at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Dr. Dönertaş is a postdoctoral researcher at the Leibniz Institute on Aging. She completed her BSc in Molecular Biology and Genetics and her MSc in Biological Sciences, both at METU. She completed her PhD at the University of Cambridge as an EMBL fellow. She worked as a postdoctoral researcher at the European Bioinformatics Institute (EMBL-EBI) and Max Planck Institute for Biology of Aging. She has worked on a variety of topics ranging from cheminformatics to evolutionary genomics and ancient DNA studies, but her main research focus is on ageing, age-related diseases, and anti-ageing interventions.

Abstract:
Unlike development, ageing is not thought of as a programmed process but a result of cellular and evolutionary stochastic events. Thus, comparative analysis of development and ageing periods can help understand the underlying mechanisms and characteristics of ageing. This seminar will summarise our recent work, comparing gene expression changes during postnatal development and ageing across individuals, organs, and cells. Using transcriptome datasets covering the whole lifespan, we study how the level and between-individual variability of gene expression changes with age. We first show that, in the human brain, increased heterogeneity is characteristic for ageing but not for development. Moreover, the temporal trend in gene expression during development does not necessarily continue to the ageing period, at which half of the expression trajectories are reversed. Studying this phenomenon in multiple tissues of mice, we found that these reversals are associated with tissue-specific functions and contribute to an interesting phenomenon that tissues diverge from each other during postnatal development but, during ageing, tend to converge towards similar expression levels. Lastly, using an external single-cell gene expression dataset, we study how tissue composition and cell-autonomous changes may contribute to this divergence-convergence pattern. Overall, our results highlight the loss of tissue- and potentially, cellular identity as a common aspect of ageing.

You can join the seminar on the Zoom platform through this link.

Can Fırtına will give a seminar on “Enabling Accurate, Fast, and Memory-Efficient Genome Analysis via Efficient and Intelligent Algorithms” on 10 January at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Can Fırtına is a Ph.D. student at ETH Zurich working with Prof. Onur Mutlu. He received B.Sc. and M.Sc. degrees in Computer Engineering from Bilkent University, in 2015 and 2018, respectively. His current research interests broadly span computational biology and computer architecture topics, including correcting sequencing errors, accurately and quickly identifying sequence similarities, and hardware/software co-design for accelerating bioinformatics applications and genomic data analysis.

Abstract:
Genome sequence analysis plays a pivotal role in enabling many medical and scientific advancements in personalized medicine, outbreak tracing, the understanding of evolution, and forensics. Modern genome sequencing technologies can rapidly generate massive amounts of genomics data at low cost. However, due to the current limitations of sequencing technologies, the analysis of genome sequencing data is currently bottlenecked by the computational and spatial overheads of existing systems and algorithms, as many of the steps in genome sequence analysis must process a large amount of data using computation. Moreover, as sequencing technologies advance, the growth in the rate that sequencing devices generate genomics data is far outpacing the corresponding growth in computational power, placing greater pressure on these bottlenecks. In this seminar, we provide an overview of our works in three main directions.

Assoc. Prof. Nurcan Tuncbag will give a seminar on “Discovery of Latent Drivers from Double Mutations in Pan-Cancer Data Reveal their Clinical Impact” on 3 January at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Dr. Tuncbag is an associate professor at Koç University jointly in the Department of Chemical and Biological Engineering and School of Medicine. She received her undergraduate degree in Chemical Engineering from Istanbul Technical University, and her master and doctorate degrees from Koç University Computational Sciences and Engineering department. She did her post-doctoral research in Massachusetts Institute of Technology (MIT) Biological Engineering Department between 2010-2014. She carried out her academic studies at Middle East Technical University between 2014-2021. Her work in computational systems biology and bioinformatics has been awarded nationally and internationally by several academies, foundations and councils. She has been a member of the Global Young Academy since 2020. She develops computational approaches to serve in network medicine, network modeling, trans-omic data integration, single cell and bulk omic data analysis, and discovery of latent cancer driver mutations.

Abstract:
Despite massive advancements in cancer genomics, to date driver mutations whose frequencies are low, and their observable translational potential is minor have escaped identification. Yet, when paired with other mutations in cis, such ‘latent driver’ mutations can drive cancer. Additionally, the additivity of co-occurring driver mutations in different genes (in trans) can lead to powerful oncogenic signal, encoding aggressive proliferation. We applied a statistical approach to identify significantly co-occurring mutations in the pan-cancer data of mutation profiles of ~80,000 tumor sequences from the TCGA and AACR GENIE databases. Evaluation of the response of cell lines and patient-derived xenograft data to drug treatment indicate that in certain genes double mutations can increase oncogenic activity, hence obtain a better drug response (e.g., in PIK3CA), or they can promote resistance to the drugs (e.g., in EGFR). On the other hand, co-occurring double mutations on different genes can additively promote tumorigenesis through single or multiple pathways. They are mostly in primary tumors. Rare occurrences can be a signature of metastatic tumors. Interrogation of big genomic data and integration with large-scale small-molecule sensitivity data can provide deep patterns that are rare – but can prompt dramatic phenotypic alterations and serve as clinical signatures.

Dr. Gökçe Ertürkmen will give a seminar on “Integrated Care Approach for the Management of Chronic Diseases in Europe and Turkey” on 27 December at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Dr. Ertürkmen has obtained her BSc, MSc, PhD from the Computer Engineering Department of the Middle East Technical University. She has finalized her PhD study on Intelligent Healthcare Monitoring Systems based on Semantically Enriched Clinical Guidelines in June 2008. She has worked as the principal researcher in many EU funded R&D projects. She has coordinated ICT-287800 SALUS project, which addresses standards-based semantic interoperability for secondary use of EHRs in pharmacovigilance domain. She has acted as the technical co-chair of the IHE QRPH domain, which is a standardization initiative in the field of quality reporting, secondary use of EHRs for research purposes and public health domain. She has published more than 90 papers in refereed international conferences and journals. Currently she is acting as the R&D director of SRDC A.Ş.

Abstract:
In this talk, Dr. Ertürkmen will present the challenges of chronic disease management, and will present an ICT infrastructure facilitating integrated care enabling multi-disciplinary care team members to collaboratively manage the care of chronic disease patients. The underlying interoperability architecture and clinical decision support systems will be introduced. Two example systems from Europe and Turkey will be presented.

Dr. Oznur Tastan will give a seminar on “Identifying Cross-Cancer Similar Patients via a Semi-Supervised Deep Clustering Approach” on 20 December at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Since 2018, Dr. Oznur Tastan has been with the Sabanci University Computer Science and Engineering, and Molecular Biology, Genetics, and Bioengineering departments. Before joining Sabanci, she worked as a faculty member at Bilkent University, Department of Computer Engineering, a post-doctoral researcher at Microsoft Research New England Lab (Cambridge, MA, USA). Dr. Öznur Taştan holds a BSc in Biological Sciences and Bioengineering from Sabanci University and received her MSc and her Ph.D. from Carnegie Mellon University, School of Computer Science, Language Technologies Institute. She has worked on diverse problems in computational biology; her present efforts center on building machine learning models to advance the current understanding of complex diseases. She is a recipient of the Young Scientist Research Award of the Science Academy (BAGEP), Tubitak Career Award, the UNESCO-L’OREAL National Fellowship for Young Women in Life Sciences, and METU Prof. Mustafa Parlar Foundation Research Incentive Award.

Abstract:
With the characterization of cancer tumors at the molecular level, there have been reports of patients being similar despite being diagnosed with different cancer types. Motivated from these observations, we aim at discovering cross-cancer patients, which we define as patients whose tumors are more similar to patient tumors diagnosed with another cancer type. We develop DeepCrossCancer to identify cross-cancer patients that always co-cluster with the other patient from another cancer type. The input to DeepCrossCancer is the transcriptomic profiles of the patient tumors, the age, and the sex of the patient. To solve the clustering problem, we use a semi-supervised deep learning-based clustering method in which the clustering task is supervised by cancer type labels and the survival times of the patients. Applying the method to patient data from nine different cancers, we discover 20 cross-cancer patients that consistently co-cluster. By analyzing the predictive genes of the cross-cancer patients and other genomic information available for the patient such as somatic mutations and copy number variations, we identify striking genomic similarities across these patients providing support. The detection of cross-cancer patients opens up possibilities for transferring clinical decisions across patients at a single patient level.

Dr. Ahmet Rifaioglu will give a seminar on “Analysis of single-cell RNA sequencing data” on 13 December at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Dr. Rifaioglu received his Ph.D. from the METU Computer Engineering department in 2020. In his Ph.D., he mainly worked on developing deep learning methods for drug discovery in the context of a project called "CROssBAR: comprehensive resource of biomedical relations with knowledge graph representations". He spent one year of his Ph.D. at EMBL-European Bioinformatics Institute as a pre-doctoral visiting researcher in the context of the CROssBAR project. After working as an assistant professor at Iskenderun Technical University for a short time, he joined Heidelberg University as a post-doctoral researcher. His research now focuses on developing computational methods for single-cell genomics in cancer.

Abstract:
With the advance of single-cell RNA sequencing technologies, we can now get whole transcriptome information for a large number of cells. These technologies enabled us to identify cell types, understand cell-cell communication and their microenvironment in general. Computational methods are required to understand and process the high-throughput single-cell transcriptomics data. In this presentation, I will talk about the standard practices and challenges in the computational analysis of single-cell RNA sequence datasets.

Asst. Prof. Aybar Can Acar will give a seminar on “Probabilistic Programming and Bayesian Inference in Biomedicine” on 6 December at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Aybar C. Acar is a faculty member at the Middle East Technical University (METU) Graduate School of Informatics. He received his BS ans MS degrees in Chemical Engineering from METU, and his PhD in Computer Science from George Mason University. His primary areas of interest are probabilistic modeling, machine learning, transcriptomics and systems biology. He is currently the co-director of the Cancer Systems Biology Laboratory at METU.

Abstract:
Probabilistic programming is a relatively new approach to problem solving where we define arbitrarily large and sophisticated probabilistic models as generative programs. We can then infer latent variables of interest by sampling from the outputs of these programs, allowing us to solve complex inference problems without having to `invert` them first. This, combined with Bayesian inference, results in a powerful way of reasoning about -- and testing hypotheses on -- poorly understood phenomena, especially under uncertainty. In this talk, I will give an overview of probabilistic programming and its evolution over the last five years, the technical difficulties involved (such as time costs of sampling) and possible remedies (e.g. variational methods and parallelization). I will also discuss the applications of this methodology to biomedical problems, with specific examples on gene expression analysis, prediction of aortic aneurysm growth, and COVID-19 epidemic projection.

Prof. Volkan Atalay will give a seminar on “Machine Learning Applications in Bioinformatics” on 29 November at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
After graduating from Ankara Atatürk Anadolu Lisesi, Volkan Atalay obtained B.S. and M.S. in Electrical and Electronics Engineering from the Middle East Technical University (METU). He received his PhD in Computer Science from Université Descartes, Paris, France in 1993. He spent a year at the New Jersey Institute of Technology, USA as a visiting scholar and another year at the Virginia Bioinformatics Institute, Virginia Tech, USA as a visiting faculty member. He is a Professor in the Department of Computer Engineering, METU, where he has been since 1993. He served as Department Chair, Assistant to President and Vice President at METU. From 2010 to 2016 he was the Chair of Board of Directors of ODTÜ TEKNOKENT (Technology Park of METU). He served as a member of Arçelik R&D Advisory Board and Board of Governors of İzmir Biomedicine and Genome Center. He was the recipient of Parlar Foundation Research Incentive Award and METU Graduate School of Natural and Applied Sciences the Best Thesis Award as the supervisor. His main research interests include machine learning applications in bioinformatics, data stream analysis and machine learning applications.

Abstract:
Bioinformatics is an interdisciplinary field that develops methods and software tools for understanding, analyzing, and interpreting biological data while machine learning deals with the development of algorithms that learn how to make predictions based on data. Thanks to the advances in computing and the growth of available data, deep learning algorithms, have been successfully applied in several fields including bioinformatics and cheminformatics. Our research group (https://github.com/cansyl) focuses on developing and applying computational techniques for the analysis of biological data and modeling of biological processes at the molecular level. We develop tools, we participate in challenges and of course we publish articles. Our effort is concentrated in automated protein function prediction and virtual screening for early drug discovery. In the first topic, we have developed DEEPred which is a hierarchical stack of multi-task feed-forward DNN and ECPred for Enzyme Commission (EC) number prediction. In virtual screening for early drug discovery, in addition to drug target interaction (DTI) prediction, we are interested in the representation of compounds and proteins and transfer learning and low-shot learning. To this end, we have developed DEEPScreen (a large-scale DTI prediction system using deep CNN), MDeePred (a large-scale DTI regression system using deep neural networks, iBioProVis (an interactive tool which embeds and visualizes compounds on 2D space) and CROssBAR (a service to generate knowledge graphs.

Prof. Tolga Can will give a seminar on “Alternative Polyadenylation (APA) events as biomarkers in cancer” on 22 November at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Dr. Tolga Can is currently a faculty member in the Department of Computer Engineering, Middle East Technical University. He received his M.Sc. and Ph.D. degrees in Computer Science from the University of California at Santa Barbara. His research interests are computational systems biology, biological networks, and graph theory. Specifically, he uses/develops bioinformatics approaches for cancer research.

Abstract:
Alternative polyadenylation (APA), a hidden complexity in cancer transcriptomes, is a process observed in recent cancer studies. In this talk, I will describe the tools we developed in collaboration with Erson Lab of METU Biological Sciences for the identification of APA events. I will also report some of the results we obtained using these tools. Specifically, we analyzed publicly available expression data for 1045 cancer patients and found a significant shift in usage of poly(A) signals in common tumor types (breast, colon, lung, prostate, gastric, and ovarian) compared to normal tissues. Using standard machine-learning techniques, we further defined specific subsets of APA events to efficiently classify cancer types. Overall, our study offers a computational approach for use of APA in novel gene discovery and classification in common tumor types, with important implications in basic research, biomarker discovery, and precision medicine approaches.

Assoc. Prof. Tunca Doğan will give a seminar on “Development and Application of Data-Driven Approaches to Make Sense out of Large-Scale and Heterogeneous Biomolecular Data” on 15 November at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Dr. Tunca Doğan has graduated from both the undergraduate and MSc programs in Middle East Technical University (METU), Faculty of Engineering, studying in the field of process engineering. He started conducting research in computational biology and bioinformatics areas during his Ph.D. study back in 2010. He received his joint Ph.D. degree from the interdisciplinary Bioengineering program, hosted by the Electrical & Electronics Engineering Department in Izmir Institute of Technology, and the Graduate School of Health Sciences and the Faculty of Medicine, DEU in 2013. Between the years 2013 and 2016, Dr. Dogan served as a post-doctoral researcher at the European Bioinformatics Institute (EMBL-EBI), Protein Function Development (UniProt) team, and at the University of Cambridge, UK. Between 2016 and 2019, he worked both as a research associate at EMBL-EBI and as an adjunct faculty member at the Institute of Informatics in METU, Turkey. Dr. Dogan currently is a faculty member at the Department of Computer Engineering, Hacettepe University, Turkey. His research interests lie in the fields of bioinformatics and cheminformatics, and can be summarized as developing and applying data science and AI-based computational methods for biomolecular function/property prediction, and computational drug discovery.

Abstract:
The recent availability of inexpensive technologies led to a surge of biological/biomedical data production and accumulation in public servers. These noisy, complex and large-scale data should be analyzed in order to understand mechanisms that constitute life and to develop new and effective treatments against prevalent diseases. A key concept in this endeavour is the prediction of unknown attributes and properties of biomolecules (i.e., genes, proteins and RNAs) such as their molecular functions, physical interactions, etc., together with their relationships to high-level biomedical concepts such as systems and diseases. Lately, cutting-edge data-driven approaches are starting to be applied to biological data to aid the development of novel and effective in silico solutions. In this seminar, I’ll summarize our efforts for integrating and representing heterogeneous data from different biological/biomedical data resources (i.e., the CROssBAR project) together with the development and application of deep learning-based computational methods for enriching the integrated data by predicting unknown functions and drug discovery centric ligand interactions of human genes and proteins.