Fig. 3From: CysPresso: a classification model utilizing deep learning protein representations to predict recombinant expression of cysteine-dense peptidesMachine learning architecture diagram of CysPresso. The primary sequence of the CDP is used to generate MSA, single, pair, and structure module AlphaFold2 representations. The four representations are concatenated, and a time series transformation utilizing random convolutional kernels is carried out on the concatenated representation. The transformed representation is then used to predict expressibility using L2-regularized logistic regression machine learning models for knottin and non-knottin CDPsBack to article page