- Browse by Subject
Browsing by Subject "translation"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
Item Global Translation of Classification Models(MDPI, 2022-05-11) Al-Merri, Mohammad; Ben Miled, Zina; Electrical and Computer Engineering, School of Engineering and TechnologyThe widespread and growing usage of machine learning models, particularly for critical areas such as law, predicate the need for global interpretability. Models that cannot be audited are vulnerable to biases inherited from the datasets that were used to develop them. Moreover, locally interpretable models are vulnerable to adversarial attacks. To address this issue, the present paper proposes a new methodology that can translate any existing machine learning model into a globally interpretable one. MTRE-PAN is a hybrid SVM-decision tree architecture that leverages the interpretability of linear hyperplanes by creating a set of polygons that delimit the decision boundaries of the target model. Moreover, the present paper introduces two new metrics: certain and boundary model parities. These metrics can be used to accurately evaluate the performance of the interpretable model near the decision boundaries. These metrics are used to compare MTRE-PAN to a previously proposed interpretable architecture called TRE-PAN. As in the case of TRE-PAN, MTRE-PAN aims at providing global interpretability. The comparisons are performed over target models developed using three benchmark datasets: Abalone, Census and Diabetes data. The results show that MTRE-PAN generates interpretable models that have a lower number of leaves and a higher agreement with the target models, especially around the most important regions in the feature space, namely the decision boundaries.Item Human keratinocyte differentiation requires translational control by the eIF2α kinase GCN2(Elsevier, 2017) Collier, Ann E.; Wek, Ronald C.; Spandau, Dan F.; Department of Dermatology, IU School of MedicineAppropriate and sequential differentiation of keratinocytes is essential for all functions of the human epidermis. While transcriptional regulation has proven to be important for keratinocyte differentiation, little is known about the role of translational control. A key mechanism for modulating translation is through phosphorylation of the α subunit of eIF2. A family of different eIF2α kinases function in the integrative stress response to inhibit general protein synthesis coincident with preferential translation of select mRNAs that participate in stress alleviation. Here we demonstrate that translational control through eIF2α phosphorylation is required for normal keratinocyte differentiation. Analyses of polysome profiles revealed that key differentiation genes, including involucrin, are bound to heavy polysomes during differentiation, despite decreased general protein synthesis. Induced eIF2α phosphorylation by the GCN2 protein kinase facilitated translational control and differentiation-specific protein expression during keratinocyte differentiation. Furthermore, loss of GCN2 thwarted translational control, normal epidermal differentiation, and differentiation gene expression in organotypic skin culture. These findings underscore a previously unknown function for GCN2 phosphorylation of eIF2α and translational control in the formation of an intact human epidermis.Item Integration of general amino acid control and TOR regulatory pathways in yeast(2010-05) Staschke, Kirk Alan; Wek, Ronald C.; Edenberg, Howard J.; Roach, Peter J.; Bard, MartinTwo important nutrient sensing and regulatory pathways, the general amino acid control (GAAC) and the target of rapamycin (TOR), participate in the control of yeast growth and metabolism in response to changes in nutrient availability. Starvation for amino acids activates the GAAC through Gcn2p phosphorylation of the translation initiation factor eIF2 and preferential translation of GCN4, a transcription activator. TOR senses nitrogen availability and regulates transcription factors, such as Gln3p. We used microarray analyses to address the integration of the GAAC and TOR pathways in directing the yeast transcriptome during amino acid starvation and rapamycin treatment. We found that the GAAC is a major effector of the TOR pathway, with Gcn4p and Gln3p each inducing a similar number of genes during rapamycin treatment. While Gcn4p activates a common core of 57 genes, the GAAC directs significant variations in the transcriptome during different stresses. In addition to inducing amino acid biosynthetic genes, Gcn4p activates genes required for assimilation of secondary nitrogen sources, such as -amino-butyric acid (GABA). Gcn2p activation upon shifting to secondary nitrogen sources is suggested to occur by means of a dual mechanism. First, Gcn2p is induced by the release of TOR repression through a mechanism involving Sit4p protein phosphatase. Second, this eIF2 kinase is activated by select uncharged tRNAs, which were shown to accumulate during the shift to GABA medium. This study highlights the mechanisms by which the GAAC and TOR pathways are integrated to recognize changing nitrogen availability and direct the transcriptome for optimal growth adaptation.Item Review: “Regina Galasso & Evelyn Scaramella, eds. Avenues of Translation: The City in Iberian and Latin American Writing. Bucknell UP, 2019(U of Kansas P, 2019-10) Mallorquí-Ruscalleda, Enric