Barros, PedroPesquita, Cátia, 1980-Rodrigues, Hugo Miguel Costa2023-05-092023-05-0920232022http://hdl.handle.net/10451/57384Tese de mestrado, Bioinformática e Biologia Computacional, 2023, Universidade de Lisboa, Faculdade de CiênciasCork oak (Quercus suber) is a mediterranean tree that excels in phellem (cork) production, a valuable raw product with multiple industrial applications. However, increased frequency and severity of drought events, due to climate change, lead to reduced cork oak growth and productivity. This work aimed to integrate different transcriptomics data available for this species to predict a gene co-expression network, and identify candidate regulatory modules of phellem development and assess their regulation by drought. The co-expression network was built using as guides a group of genes differentially expressed in phellem from plants exposed to drought conditions. Based on gene-to-gene co-expression links, transcription factor (TF)-target gene interactions were further predicted and reinforced using functional data available in the model plant A. thaliana. The generated network highlighted predominantly genes negatively regulated under drought, particularly gene modules related to cell division and differentiation (e.g. cell wall development). From the multiple interactions established by co-expression involving 27 TFs, 118 had been already identified in A. thaliana by experimental methods. Additionally, the specific binding sites predicted for 21 TFs were found in the promoters of 144 co-expressed genes. This demonstrated that the predicted co-expression network could, to some extent, predict candidate TF-target interactions. From the highlighted TFs, MYB93 and NAC43 were, respectively, integrated in two modules showing a concerted downregulation in phellem, in response to drought. Additionally, DREB1B TF and co-expressed targets are hypothesized as being involved in a adaptive response to drought by maintaining cellular homeostasis. Overall, the present work unveiled new regulatory gene modules of interest using stateof- the-art machine learning and data mining approaches, with predicted role in phellem development, and described their expression trend in response to drought. The selected candidate transcription factors will be further experimentally validated to reinforce the obtained in silico predictions. This will be an important contribution for the development of future strategies to screen cork oak plants for improved resilience and/or productivity, in response to adverse external conditions.engSobreironetwork de co-expressãosecuramineração de dadostranscriptomaTeses de mestrado - 2023master thesis203485637