NeuroChaos Learning (NL) is a chaos-based learning algorithm inspired by the chaotic firing of neurons in the human brain. This method suggests a combination of traditional machine learning algorithms with feature transformation, neurochaos feature extraction, and modifications based on chaotic maps. The architecture used here can be viewed as an application of chaos as a kernel trick applicable to various machine learning algorithms. In this work, we use NL to predict whether a given protein is a hypothetical protein [HP], a significant area of research in bioinformatics. The F1-scores of conventional machine learning techniques, ChaosNet, and the NL algorithms, with noticeably less training data, are compared.