3D-QSARpy: Combining variable selection strategies and machine learning techniques to build QSAR models

Abstract Quantitative Structure-Activity Relationship (QSAR) is a computer-aided technology in the field of medicinal chemistry that seeks to clarify the relationships between molecular structures and their biological activities. Such technologies allow for the acceleration of the development of new compounds by reducing the costs of drug design. This work presents 3D-QSARpy, a flexible, user-friendly and robust tool, freely available without registration, to support the generation of QSAR 3D models in an automated way. The user only needs to provide aligned molecular structures and the respective dependent variable. The current version was developed using Python with packages such as scikit-learn and includes various techniques of machine learning for regression. The diverse techniques employed by the tool is a differential compared to known methodologies, such as CoMFA and CoMSIA, because it expands the search space of possible solutions, and in this way increases the chances of obtaining relevant models. Additionally, approaches for select variables (dimension reduction) were implemented in the tool. To evaluate its potentials, experiments were carried out to compare results obtained from the proposed 3D-QSARpy tool with the results from already published works. The results demonstrated that 3D-QSARpy is extremely useful in the field due to its expressive results.

Advancements in artificial intelligence and machine learning in revolutionising biomarker discovery

Abstract The article explores the significance of biomarkers in clinical research and the advantages of utilizing artificial intelligence (AI) and machine learning (ML) in the discovery process. Biomarkers provide a more comprehensive understanding of disease progression and response to therapy compared to traditional indicators. AI and ML offer a new approach to biomarker discovery, leveraging large amounts of data to identify patterns and optimize existing biomarkers. Additionally, the article touches on the emergence of digital biomarkers, which use technology to assess an individual's physiological and behavioural states, and the importance of properly processing omics and multi-omics data for efficient handling by computer systems. However, the article acknowledges the challenges posed by AI/ML in the identification of biomarkers, including potential biases in the data and the need for diversity in data representation. To address these challenges, the article suggests the importance of regulation and diversity in the development of AI/ML algorithms.

A map of neutrophil granules by proteomics / Mapa dos neutrófilos por proteômica (OU) Mapa dos grânulos do neutrófilo por proteômica

Neutrophils are polymorphonuclear leukocytes that play a key role in the organism defense. These cells enroll in a range of actions to ensure pathogen elimination and orchestrate both innate and adaptative immune responses. The main physiological structures of neutrophils are their storage organelles that are essential since the cells activation and participate in all their functions. The storage organelles are divided into 2 types: granules and secretory vesicles. The granules are subdivided into azurophilic, specific and gelatinase. The granules are distinguished by their protein content, and since they play an important role on the neutrophil function, the knowledge of the proteins stored in these organelles can help to better understand these cells. Some proteins are present in high abundance and are used as markers for each storage organelle. These proteins are myeloperoxidase (MPO) for azurophil granules, neutrophil gelatinase associated with lipocalin-2 (NGAL) and lactoferrin (LTF) for specific granules, matrix metalloproteinase-9 (MMP9) for gelatinase granules and alkaline phosphatase (AP) for secretory vesicles. The isolation of neutrophils granules, however, is challenging and the existing procedures rely on large sample volumes, about 400 mL of peripheral blood or 3 x 108 neutrophils, not allowing for multiple biological and technical replicates. Therefore, the aim of this study was to develop a miniaturized neutrophil granules isolation method and to use biochemical assays, mass spectrometry-based proteomics and a machine learning approach to investigate the protein content of the neutrophils storage organelles. With that in mind, 40 mL of the peripheral blood of three apparently healthy volunteers were collected. The neutrophils were isolated, disrupted using nitrogen cavitation and organelles were fractionated with a discontinuous 3-layer Percoll density gradient. The presence of granules markers in each fraction was assessed using western blot , gelatin zymography and enzymatic assays. The isolation was proven successful and allowed for a reasonable separation of all neutrophils storage organelles in a gradient of less than 1 mL, about 37 times smaller than the methodsdescribed in the literature. Moreover, mass spectrometry-based proteomics identified 369 proteins in at least 3 of the 5 samples, and using a machine learning strategy, the localization of 140 proteins was predicted with confidence. Furthermore, this study was the first to investigate the proteome of neutrophil granules using technical and biological replicates, creating a reliable database for further studies. In conclusion, the developed miniaturized method is reproducible, cheaper, and reliable. In addition, it provides a resource for further studies exploring neutrophil granules protein content and mobilization during activation with different stimuli

Neutrófilos são leucócitos polimorfonucleares que possuem papel fundamental na defesa do organismo. Essas células desempenham diversas ações a fim de assegurar a eliminação de um patógeno e, além disso, orquestram a resposta imune inata e adaptativa. O conjunto composto pelos grânulos de armazenamento e as vesículas secretórias compõe a principal estrutura fisiológica dos neutrófilos. Estes componentes são essenciais desde a ativação celular, participando de todas as funcionalidades desta célula. Os grânulos são subdivididos em azurófilos, específicos e gelatinase. Eles podem ser distinguidos por meio de seu conteúdo proteico e, como são importantes na funcionalidade dos neutrófilos, identificar quais proteínas são armazenadas nestas organelas é imprescindível para entender melhor essa célula como um todo. Algumas proteínas, estão presentes de forma abundante e, portanto, são utilizadas como marcadores dos grânulos. Tais proteínas são mieloperoxidase (MPO) para os grânulos azurófilos, gelatinase de neutrófilo associada a lipocalina (NGAL) e lactoferrina (LTF) para os específicos, metaloproteinase de matrix 9 (MMP9) para os grânulos de gelatinase e fosfatase alcalina (AP) para as vesículas secretórias. Isolar estas estruturas, no entanto, é desafiador visto que os protocolos existentes na literatura utilizam grandes volumes de amostra, cerca de 400 mL de sangue ou 3 x 108 neutrófilos, para apenas um isolamento, impedindo a realização de replicatas técnicas e biológicas. Desta forma, o objetivo do presente estudo foi desenvolver um protocolo miniaturizado de isolamento dos grânulos neutrofílicos e utilizar métodos bioquímicos, de proteômica e machine learning para investigar o conteúdo proteico destas estruturas celulares. Para isto, 40 mL de sangue periférico de três voluntários aparentemente saudáveis foi coletado. Os neutrófilos foram então isolados, lisados com cavitação de nitrogênio e o fracionamento subcelular foi realizado baseado em um gradiente descontínuo de 3 camadas de Percoll. O método de isolamento foi avaliado através da investigação dos marcadores utilizando western blotting (WB), zimografia de gelatina e ensaios enzimáticos em cada fração coletada. O isolamento demonstrou-se eficiente e permitiu uma ótima separação dos grânulosem um gradiente menor que 1 mL, cerca de 37 vezes menor que os métodos atualmente descritos na literatura. Além disso, a análise proteômica foi capaz de identificar 369 proteínas presentes em pelo menos 3 das 5 réplicas investigadas e, utilizando ferramentas de machine learning, 140 proteínas foram classificadas como pertencentes a um dos tipos de grânulos ou vesícula secretória com alto nível de confiabilidade. Por fim, o presente estudo foi o primeiro a investigar o proteoma dos grânulos utilizando replicatas técnicas e biológicas, criando e fornecendo uma base de dados robusta que poderá ser utilizada em estudos futuros. Conclui-se, portanto, que a metodologia miniaturizada desenvolvida é eficaz, reprodutível e mais barata, além de permitir estudos mais complexos e profundos sobre o proteoma dos grânulos dos neutrófilos em diferentes momentos celulares, tais como quando ativados via estímulos distintos

Identification of autism spectrum disorder using deep learning and the ABIDE dataset.

The goal of the present study was to apply deep learning algorithms to identify autism spectrum disorder (ASD) patients from large brain imaging dataset, based solely on the patients brain activation patterns. We investigated ASD patients brain imaging data from a world-wide multi-site database known as ABIDE (Autism Brain Imaging Data Exchange). ASD is a brain-based disorder characterized by social deficits and repetitive behaviors. According to recent Centers for Disease Control data, ASD affects one in 68 children in the United States. We investigated patterns of functional connectivity that objectively identify ASD participants from functional brain imaging data, and attempted to unveil the neural patterns that emerged from the classification. The results improved the state-of-the-art by achieving 70% accuracy in identification of ASD versus control patients in the dataset. The patterns that emerged from the classification show an anticorrelation of brain function between anterior and posterior areas of the brain; the anticorrelation corroborates current empirical evidence of anterior-posterior disruption in brain connectivity in ASD. We present the results and identify the areas of the brain that contributed most to differentiating ASD from typically developing controls as per our deep learning model.