Dhiraj Pyakurel

Accurate prediction of Major Histocompatibility Complex (MHC) class II-peptide binding affinity remains a critical challenge in immunotherapy and vaccine development due to the complexity of molecular interactions and allele-specific binding preferences. While existing computational approaches have shown promise, identifying the optimal deep learning architecture for this task remains an open question. This study addresses this gap by systematically comparing three state-of-the-art deep learning architectures: Long Short-Term Memory Networks (LSTM), Transformers, and Temporal Convolutional Networks (TCN) using the IEDB2016 dataset containing 134,281 MHC class II-peptide binding records. Our comprehensive evaluation employed five-fold cross-validation and ensemble modeling to assess both regression and classification performance. Results demonstrate that TCN consistently outperforms competing architectures, achieving superior regression metrics with R² of 0.6208 (vs. LSTM: 0.5923, Transformer: 0.5706) and classification performance with AUROC of 0.8766 (vs. LSTM: 0.8736, Transformer: 0.8707). Cross-validation confirmed TCN's robustness (average R²: 0.6229), while ensemble methods further validated its superiority (R²: 0.6805). These findings establish TCN as the most effective architecture for peptide-MHC binding prediction, offering significant implications for computational immunology and therapeutic design.

Recognizing handwritten text plays a crucial role in converting scanned documents, whether printed or handwritten, into editable and searchable formats. In this study, various models such as CRNN, TCN, and Transformer have been utilized for Handwritten Text Recognition (HTR), where input data consists of sequences of image patches representing English text. The CRNN model employed comprises three layers: a CNN for extracting feature maps from handwritten text images, and Bidirectional Long-Short Term Memory (BLSTM) and Bidirectional Gated Recurrent Unit (BiGRU) in the RNN layer to address the gradient vanishing/exploding issue of simple RNNs. Additionally, TCN and Transformer models are employed for HTR. Optimizers including SGD, RMSprop, Adam, and Adamax, along with fine-tuning of hyperparameters, are utilized to enhance model accuracy. Model performance is evaluated using metrics such as f1 Score, precision, and recall. Meta learner optimization is subsequently employed to enhance the performance of deep learning models. The IAM dataset in English is utilized for training, validation, and testing. The Bi-LSTM model achieves an accuracy of 90.04%, precision of 91.62%, recall of 88.98%, and an f1 Score of 0.9025. With TCN, similar metrics are achieved. The Transformer model achieves an accuracy of 85.86%, precision of 88.94%, recall of 83.86%, and an f1 Score of 0.8626. Furthermore, Bi-GRU achieves an accuracy of 90.32%, precision of 91.56%, recall of 89.53%, and an f1 Score of 0.9050. Following the basic models, a meta model is constructed for the best performing model, demonstrating significant enhancement in Handwritten Text Recognition with an accuracy of 92.30%, precision of 94.80%, recall of 93.18%, and an f1 Score of 0.9425.