Antibiotic resistance poses a global health issue that requires innovative techniques for predicting antibiotic resistance genes (ARGs). Traditionally, prediction models depended on sequence-based methods that examined protein sequences to detect ARGs. However, these approaches frequently encounter challenges in identifying the connections between proteins and their interactions. To address this problem, we introduce GNN-ARG, a framework that investigates graph neural network (GNN) architectures to predict ARG from protein sequences using protein interaction graphs. Our method builds a weighted, undirected graph where nodes correspond to proteins and links indicate how similar they are. We utilize existing ESM-2 embeddings as features for nodes to capture specific sequence details. For training GNN-ARG, we employed a dataset combining protein sequences categorized as either ARG or non-ARG from nine public ARG databases. The model was trained with the aim of predicting node labels through binary classification tasks. In the GNN-ARG framework, we conducted a thorough assessment of four well-known GNN models: Graph Convolutional Networks (GCN), Graph Isomorphism Networks (GIN), Graph Attention Networks (GAT), and Graph Sage. Our experimental findings show that GIN outperforms the others providing an accuracy of 93.22% with an F1 score of 0.9161, followed by GCN and Graph Sage, which also show performance closely behind GIN. Despite providing insights into node importance, the GAT model achieves both lower accuracy and F1 score. These results highlight how GNN models can improve the prediction of resistance genes and help us gain a deeper understanding of ways to address antibiotic resistance more effectively in the future research field of bioinformatics by exploring graph-based methods extensively.

Identifying out-of-distribution (OOD) data at inference time is crucial for many machine learning applications, especially for automation. We present a novel unsupervised semi-parametric framework COMBOOD for OOD detection with respect to image recognition. Our framework combines signals from two distance metrics, nearest-neighbor and Mahalanobis, to derive a confidence score for an inference point to be out-of-distribution. The former provides a non-parametric approach to OOD detection. The latter provides a parametric, simple, yet effective method for detecting OOD data points, especially, in the far OOD scenario, where the inference point is far apart from the training data set in the embedding space. However, its performance is not satisfactory in the near OOD scenarios that arise in practical situations. Our COMBOOD framework combines the two signals in a semi-parametric setting to provide a confidence score that is accurate both for the near-OOD and far-OOD scenarios. We show experimental results with the COMBOOD framework for different types of feature extraction strategies. We demonstrate experimentally that COMBOOD outperforms state-of-the-art OOD detection methods on the OpenOOD (both version 1 and most recent version 1.5) benchmark datasets (for both far-OOD and near-OOD) as well as on the documents dataset in terms of accuracy.

Satellite image classification is a challenging problem that lies at the crossroads of remote sensing, computer vision, and machine learning. Due to the high variability inherent in satellite data, most of the current object classification approaches are not suitable for handling satellite datasets. The progress of satellite image analytics has also been inhibited by the lack of a single labeled high-resolution dataset with multiple class labels. The contributions of this paper are twofold -- (1) first, we present two new satellite datasets called SAT-4 and SAT-6, and (2) then, we propose a classification framework that extracts features from an input image, normalizes them and feeds the normalized feature vectors to a Deep Belief Network for classification. On the SAT-4 dataset, our best network produces a classification accuracy of 97.95% and outperforms three state-of-the-art object recognition algorithms, namely - Deep Belief Networks, Convolutional Neural Networks and Stacked Denoising Autoencoders by ~11%. On SAT-6, it produces a classification accuracy of 93.9% and outperforms the other algorithms by ~15%. Comparative studies with a Random Forest classifier show the advantage of an unsupervised learning approach over traditional supervised learning techniques. A statistical analysis based on Distribution Separability Criterion and Intrinsic Dimensionality Estimation substantiates the effectiveness of our approach in learning better representations for satellite imagery.