Smartpathk: a platform for teaching glomerulopathies using machine learning.

BACKGROUND: With the emergence of the new coronavirus pandemic (COVID-19), distance learning, especially that mediated by information and digital communication technologies, has been adopted in all areas of knowledge and at all levels, including medical education. Imminently practical areas, such as pathology, have made traditional teaching based on conventional microscopy more flexible through the synergies of computational tools and image digitization, not only to improve teaching-learning but also to offer alternatives to repetitive and exhaustive histopathological analyzes. In this context, machine learning algorithms capable of recognizing histological patterns in kidney biopsy slides have been developed and validated with a view to building computational models capable of accurately identifying renal pathologies. In practice, the use of such algorithms can contribute to the universalization of teaching, allowing quality training even in regions where there is a lack of good nephropathologists. The purpose of this work is to describe and test the functionality of SmartPathk, a tool to support teaching of glomerulopathies using machine learning. The training for knowledge acquisition was performed automatically by machine learning methods using the J48 algorithm to create a computational model of an appropriate decision tree. RESULTS: An intelligent system, SmartPathk, was developed as a complementary remote tool in the teaching-learning process for pathology teachers and their students (undergraduate and graduate students), showing 89,47% accuracy using machine learning algorithms based on decision trees. CONCLUSION: This artificial intelligence system can assist in teaching renal pathology to increase the training capacity of new medical professionals in this area.

pHLA3D: Updating the database of predicted three-dimensional structures of HLA with HLA-DR, HLA-DQ and HLA-DP molecules.

To improve the availability of three-dimensional (3D) structures of HLA molecules, we created the pHLA3D database. In its first version, we modeled and published 106 3D structures of HLA class I molecules from the HLA-A, HLA-B, and HLA-C loci. This paper presents an update of this database, providing more 127 3D structures of HLA class II molecules (41 DR, 42 DQ, and 44 DP), predicted via homology modeling with MODELLER and SWISS-MODEL. These new 3D structures of HLA class II molecules are now freely available at pHLA3D (www.phla3d.com.br) for immunologists and other researchers working with HLA molecules.

EpHLA software: a timesaving and accurate tool for improving identification of acceptable mismatches for clinical purposes.

UNLABELLED: The HLAMatchmaker algorithm, which allows the identification of "safe" acceptable mismatches (AMMs) for recipients of solid organ and cell allografts, is rarely used in part due to the difficulty in using it in the current Excel format. The automation of this algorithm may universalize its use to benefit the allocation of allografts. Recently, we have developed a new software called EpHLA, which is the first computer program automating the use of the HLAMatchmaker algorithm. Herein, we present the experimental validation of the EpHLA program by showing the time efficiency and the quality of operation. The same results, obtained by a single antigen bead assay with sera from 10 sensitized patients waiting for kidney transplants, were analyzed either by conventional HLAMatchmaker or by automated EpHLA method. Users testing these two methods were asked to record: (i) time required for completion of the analysis (in minutes); (ii) number of eplets obtained for class I and class II HLA molecules; (iii) categorization of eplets as reactive or non-reactive based on the MFI cutoff value; and (iv) determination of AMMs based on eplets' reactivities. We showed that although both methods had similar accuracy, the automated EpHLA method was over 8 times faster in comparison to the conventional HLAMatchmaker method. In particular the EpHLA software was faster and more reliable but equally accurate as the conventional method to define AMMs for allografts. CONCLUSION: The EpHLA software is an accurate and quick method for the identification of AMMs and thus it may be a very useful tool in the decision-making process of organ allocation for highly sensitized patients as well as in many other applications.

Expression of the zinc transporters genes and metallothionein in obese women.

Research has investigated the participation of zinc transport proteins and metallothionein in the metabolism of this mineral. However, studies about the genetic expression of these proteins in obese patients are scarce. The study determined the expression of zinc transporter protein codifying genes (ZnT-1, Zip-1 and Zip-3) and of metallothionein in 55 obese women, aged between 20 and 56 years. The assessment of body composition was carried out using anthropometric measurements and bioelectrical impedance. Zinc intake was obtained by recording diet over a 3-day period, and the nutritional analysis was carried out using NutWin software version 1.5. The plasmatic and erythrocytary zinc were analyzed by atomic absorption spectrophotometry (λ=213. 9 nm). The determination of mRNA expression of the zinc transporter proteins and metallothionein was carried out using blood, using the RT-PCR method. The mean values of body mass index were 37.9±5.5 kg/m2. The average intake of zinc was 9.4±2.3 mg/day. The analysis of the zinc plasma concentrations showed values of 58.4±10.9 µg/dL. The mean values of zinc in the erythroytes were 38.7±9.1 µg/g Hb. The metallothionein gene had a higher expression in the blood, when compared to zinc transporters ZnT-1, Zip-1, and Zip-3 (p=0.01). The study shows that there are alterations in the biochemical parameters of zinc in obese patients assessed, as well as higher expression of the codifying gene metallothionein, when compared to the investigated zinc transporters.