IAV-CNN: a 2D convolutional neural network model to predict antigenic variants of influenza A virus

Abstract

The rapid evolution of influenza viruses constantly leads to the emergence of novel influenza strains that are capable of escaping from population immunity. The timely determination of antigenic variants is critical to vaccine design. Empirical experimental methods like hemagglutination inhibition (HI) assays are time-consuming and labor-intensive, requiring live viruses. Recently, many computational models have been developed to predict the antigenic variants without considerations of explicitly modeling the interdependencies between the channels of feature maps. Moreover, the influenza sequences consisting of similar distribution of residues will have high degrees of similarity and will affect the prediction outcome. Consequently, it is challenging but vital to determine the importance of different residue sites and enhance the predictive performance of influenza antigenicity. We have proposed a 2D convolutional neural network (CNN) model to infer influenza antigenic variants (IAV-CNN). Specifically, we introduce a new distributed representation of amino acids, named ProtVec that can be applied to a variety of downstream proteomic machine learning tasks. After splittings and embeddings of influenza strains, a 2D squeeze-and-excitation CNN architecture is constructed that enables networks to focus on informative residue features by fusing both spatial and channel-wise information with local receptive fields at each layer. Experimental results on three influenza datasets show IAV-CNN achieves state-of-the-art performance combing the new distributed representation with our proposed architecture. It outperforms both traditional machine algorithms with the same feature representations and the majority of existing models in the independent test data. Therefore we believe that our model can be served as a reliable and robust tool for the prediction of antigenic variants.

VaxSeer: Selecting influenza vaccines through evolutionary and antigenicity models

“…We next compared our antigenicity predictor with previous research applying traditional machine learning methods [21,22] and other deep learning models [23,24] to this problem. One representative baseline is a linear regression model, denoted as LR+ [22].…”

“…When combined, these models allow us to compute the predicted coverage score, a vaccine's expected effectiveness over a panel of potential circulating viruses. Compared to existing assays or metrics, which model individual aspects of immunology like viral fitness [14][15][16][17][18][19][20] and antigenic properties [21][22][23][24], our goal is to optimize a holistic score that prospectively quantifies the effectiveness of vaccines. Retrospective analyses show that the coverage score is strongly correlated with vaccine effectiveness and reduction of disease burden.…”

“…We leverage both of these predictors to produce an "escapability score," which is used to rank and select the optimal vaccine strains. There exist computational methods that model viral evolution [12][13][14][15][16] or antigenicity [17][18][19][20] as independent factors. However, to the best of our knowledge, we are the first to connect these aspects into a unified computational pipeline for vaccine selection.…”

“…We also compare our antigenicity predictor with previous research applying traditional machine learning methods [17,18] and other deep learning models [19,20] to this problem. We compare our model with two representative models.…”

User-specific recommender systems: from data to model

“…where * represents a convolution operator, W 𝑙 𝑖 ∈ R |N(𝑖)|×𝐻 is a weight matrix and b 𝑙 𝑖 is the corresponding bias. Besides, 𝜓(•) is a non-linear activation function, here we employ the rectified linear unit (ReLU) [67,107,108]. Note that |N(𝑖)| × 𝐻 denotes the size of the filter, aiming to cover all the rows (historical items) of the embedding map.…”