Araştırma Çıktıları
Permanent URI for this communityhttps://hdl.handle.net/11443/931
Browse
6 results
Search Results
Item A framework for validating AI in precision medicine: considerations from the European ITFoC consortium(BMC, 2021-01-01) Tsopra, Rosy; Fernandez, Xose; Luchinat, Claudio; Alberghina, Lilia; Lehrach, Hans; Vanoni, Marco; Dreher, Felix; Sezerman, O. Ugur; Cuggia, Marc; de Tayrac, Marie; Miklasevics, Edvins; Itu, Lucian Mihai; Geanta, Marius; Ogilvie, Lesley; Godey, Florence; Boldisor, Cristian Nicolae; Campillo-Gimenez, Boris; Cioroboiu, Cosmina; Ciusdel, Costin Florian; Coman, Simona; Cubelos, Oliver Hijano; Itu, Alina; Lange, Bodo; Le Gallo, Matthieu; Lespagnol, Alexandra; Mauri, Giancarlo; Soykam, H. Okan; Rance, Bastien; Turano, Paola; Tenori, Leonardo; Vignoli, Alessia; Wierling, Christoph; Benhabiles, Nora; Burgun, AnitaBackground Artificial intelligence (AI) has the potential to transform our healthcare systems significantly. New AI technologies based on machine learning approaches should play a key role in clinical decision-making in the future. However, their implementation in health care settings remains limited, mostly due to a lack of robust validation procedures. There is a need to develop reliable assessment frameworks for the clinical validation of AI. We present here an approach for assessing AI for predicting treatment response in triple-negative breast cancer (TNBC), using real-world data and molecular -omics data from clinical data warehouses and biobanks. Methods The European ``ITFoC (Information Technology for the Future Of Cancer){''} consortium designed a framework for the clinical validation of AI technologies for predicting treatment response in oncology. Results This framework is based on seven key steps specifying: (1) the intended use of AI, (2) the target population, (3) the timing of AI evaluation, (4) the datasets used for evaluation, (5) the procedures used for ensuring data safety (including data quality, privacy and security), (6) the metrics used for measuring performance, and (7) the procedures used to ensure that the AI is explainable. This framework forms the basis of a validation platform that we are building for the ``ITFoC Challenge{''}. This community-wide competition will make it possible to assess and compare AI algorithms for predicting the response to TNBC treatments with external real-world datasets. Conclusions The predictive performance and safety of AI technologies must be assessed in a robust, unbiased and transparent manner before their implementation in healthcare settings. We believe that the consideration of the ITFoC consortium will contribute to the safe transfer and implementation of AI in clinical settings, in the context of precision oncology and personalized care.Item IDH-mutant glioma specific association of rs55705857 located at 8q24.21 involves MYC deregulation(NATURE PUBLISHING GROUP, 2016-01-01) Oktay, Yavuz; Ulgen, Ege; Can, Ozge; Akyerli, Cemaliye B.; Yuksel, Sirin; Erdemgil, Yigit; Durasi, I. Melis; Henegariu, Octavian Ioan; Nanni, E. Paolo; Selevsek, Nathalie; Grossmann, Jonas; Erson-Omay, E. Zeynep; Bai, Hanwen; Gupta, Manu; Lee, William; Turcan, Sevin; Ozpinar, Aysel; Huse, Jason T.; Sav, M. Aydin; Flanagan, Adrienne; Gunel, Murat; Sezerman, O. Ugur; Yakicier, M. Cengiz; Pamir, M. Necmettin; Ozduman, KorayThe single nucleotide polymorphism rs55705857, located in a non-coding but evolutionarily conserved region at 8q24.21, is strongly associated with IDH-mutant glioma development and was suggested to be a causal variant. However, the molecular mechanism underlying this association has remained unknown. With a case control study in 285 gliomas, 316 healthy controls, 380 systemic cancers, 31 other CNS-tumors, and 120 IDH-mutant cartilaginous tumors, we identified that the association was specific to IDH-mutant gliomas. Odds-ratios were 9.25 (5.17-16.52Item Mutations and Copy Number Alterations in IDH Wild-Type Glioblastomas Are Shaped by Different Oncogenic Mechanisms(MDPI, 2020-01-01) Ulgen, Ege; Karacan, Sila; Gerlevik, Umut; Can, Ozge; Bilguvar, Kaya; Oktay, Yavuz; B. Akyerli, Cemaliye; K. Yuksel, Sirin; E. Danyeli, Ayca; Tihan, Tarik; Sezerman, O. Ugur; Yakicier, M. Cengiz; Pamir, M. Necmettin; Ozduman, KorayLittle is known about the mutational processes that shape the genetic landscape of gliomas. Numerous mutational processes leave marks on the genome in the form of mutations, copy number alterations, rearrangements or their combinations. To explore gliomagenesis, we hypothesized that gliomas with different underlying oncogenic mechanisms would have differences in the burden of various forms of these genomic alterations. This was an analysis on adult diffuse gliomas, but IDH-mutant gliomas as well as diffuse midline gliomas H3-K27M were excluded to search for the possible presence of new entities among the very heterogenous group of IDH-WT glioblastomas. The cohort was divided into two molecular subsets: (1) Molecularly-defined GBM (mGBM) as those that carried molecular features of glioblastomas (including TERT promoter mutations, 7/10 pattern, or EGFR-amplification), and (2) those who did not (others). Whole exome sequencing was performed for 37 primary tumors and matched blood samples as well as 8 recurrences. Single nucleotide variations (SNV), short insertion or deletions (indels) and copy number alterations (CNA) were quantified using 5 quantitative metrics (SNV burden, indel burden, copy number alteration frequency-wGII, chromosomal arm event ratio-CAER, copy number amplitude) as well as 4 parameters that explored underlying oncogenic mechanisms (chromothripsis, double minutes, microsatellite instability and mutational signatures). Findings were validated in the TCGA pan-glioma cohort. mGBM and ``Others{''} differed significantly in their SNV (only in the TCGA cohort) and CNA metrics but not indel burden. SNV burden increased with increasing age at diagnosis and at recurrences and was driven by mismatch repair deficiency. On the contrary, indel and CNA metrics remained stable over increasing age at diagnosis and with recurrences. Copy number alteration frequency (wGII) correlated significantly with chromothripsis while CAER and CN amplitude correlated significantly with the presence of double minutes, suggesting separate underlying mechanisms for different forms of CNA.Item CSF Proteomics Identifies Specific and Shared Pathways for Multiple Sclerosis Clinical Subtypes(PUBLIC LIBRARY SCIENCE, 2015-01-01) Avsar, Timucin; Durasi, Ilknur Melis; Uygunoglu, Ugur; Tutuncu, Melih; Demirci, Nuri Onat; Saip, Sabahattin; Sezerman, O. Ugur; Siva, Aksel; Turanli, Eda TahirMultiple sclerosis (MS) is an immune-mediated, neuro-inflammatory, demyelinating and neurodegenerative disease of the central nervous system (CNS) with a heterogeneous clinical presentation and course. There is a remarkable phenotypic heterogeneity in MS, and the molecular mechanisms underlying it remain unknown. We aimed to investigate further the etiopathogenesis related molecular pathways in subclinical types of MS using proteomic and bioinformatics approaches in cerebrospinal fluids of patients with clinically isolated syndrome, relapsing remitting MS and progressive MS (n=179). Comparison of disease groups with controls revealed a total of 151 proteins that are differentially expressed in clinically different MS subtypes. KEGG analysis using PANOGA tool revealed the disease related pathways including aldosterone-regulated sodium reabsorption (p=8.02x10(-5)) which is important in the immune cell migration, renin-angiotensin (p=6.88x10(-5)) system that induces Th17 dependent immunity, notch signaling (p=1.83x10(-10)) pathway indicating the activated remyelination and vitamin digestion and absorption pathways (p=1.73x10(-5)). An emerging theme from our studies is that whilst all MS clinical forms share common biological pathways, there are also clinical subtypes specific and pathophysiology related pathways which may have further therapeutic implications.Item Sequential filtering for clinically relevant variants as a method for clinical interpretation of whole exome sequencing findings in glioma(BMC, 2021-01-01) Ulgen, Ege; Can, Ozge; Bilguvar, Kaya; Boylu, Cemaliye Akyerli; Yuksel, Sirin Kilicturgay; Danyeli, Ayca Ersen; Sezerman, O. Ugur; Yakicier, M. Cengiz; Pamir, M. Necmettin; Ozduman, KorayBackground In the clinical setting, workflows for analyzing individual genomics data should be both comprehensive and convenient for clinical interpretation. In an effort for comprehensiveness and practicality, we attempted to create a clinical individual whole exome sequencing (WES) analysis workflow, allowing identification of genomic alterations and presentation of neurooncologically-relevant findings. Methods The analysis workflow detects germline and somatic variants and presents: (1) germline variants, (2) somatic short variants, (3) tumor mutational burden (TMB), (4) microsatellite instability (MSI), (5) somatic copy number alterations (SCNA), (6) SCNA burden, (7) loss of heterozygosity, (8) genes with double-hit, (9) mutational signatures, and (10) pathway enrichment analyses. Using the workflow, 58 WES analyses from matched blood and tumor samples of 52 patients were analyzed: 47 primary and 11 recurrent diffuse gliomas. Results The median mean read depths were 199.88 for tumor and 110.955 for normal samples. For germline variants, a median of 22 (14-33) variants per patient was reported. There was a median of 6 (0-590) reported somatic short variants per tumor. A median of 19 (0-94) broad SCNAs and a median of 6 (0-12) gene-level SCNAs were reported per tumor. The gene with the most frequent somatic short variants was TP53 (41.38\%). The most frequent chromosome-/arm-level SCNA events were chr7 amplification, chr22q loss, and chr10 loss. TMB in primary gliomas were significantly lower than in recurrent tumors (p = 0.002). MSI incidence was low (6.9\%). Conclusions We demonstrate that WES can be practically and efficiently utilized for clinical analysis of individual brain tumors. The results display that NOTATES produces clinically relevant results in a concise but exhaustive manner.Item driveR: a novel method for prioritizing cancer driver genes using somatic genomics data(BMC, 2021-01-01) Ulgen, Ege; Sezerman, O. UgurBackground: Cancer develops due to ``driver{''} alterations. Numerous approaches exist for predicting cancer drivers from cohort-scale genomics data. However, methods for personalized analysis of driver genes are underdeveloped. In this study, we developed a novel personalized/batch analysis approach for driver gene prioritization utilizing somatic genomics data, called driveR. Results: Combining genomics information and prior biological knowledge, driveR accurately prioritizes cancer driver genes via a multi-task learning model. Testing on 28 different datasets, this study demonstrates that driveR performs adequately, achieving a median AUC of 0.684 (range 0.651-0.861) on the 28 batch analysis test datasets, and a median AUC of 0.773 (range 0-1) on the 5157 personalized analysis test samples. Moreover, it outperforms existing approaches, achieving a significantly higher median AUC than all of MutSigCV (Wilcoxon rank-sum test p < 0.001), DriverNet (p < 0.001), OncodriveFML (p < 0.001) and MutPanning (p < 0.001) on batch analysis test datasets, and a significantly higher median AUC than DawnRank (p < 0.001) and PRODIGY (p < 0.001) on personalized analysis datasets. Conclusions: This study demonstrates that the proposed method is an accurate and easy-to-utilize approach for prioritizing driver genes in cancer genomes in personalized or batch analyses. driveR is available on CRAN: https://cran.r-project.org/package=driveR.