PScL-DDCFPred: an ensemble deep learning-based approach for characterizing multiclass subcellular localization of human proteins from bioimage data

Abstract

Motivation Characterization of protein subcellular localization has become an important and long-standing task in bioinformatics and computational biology, which provides valuable information for elucidating various cellular functions of proteins and guiding drug design. Results Here, we develop a novel bioimage-based computational approach, termed PScL-DDCFPred, to accurately predict protein subcellular localizations in human tissues. PScL-DDCFPred first extracts multiview image features, including global and local features, as base or pure features; Next, it applies a new integrative feature selection method based on stepwise discriminant analysis and generalized discriminant analysis to identify the optimal feature sets from the extracted pure features; Finally, a classifier based on deep neural network (DNN) and deep-cascade forest (DCF) is established. Stringent ten-fold cross-validation tests on the new protein subcellular localization training dataset, constructed from the human protein atlas databank, illustrates that PScL-DDCFPred achieves a better performance than several existing state-of-the-art methods. Moreover, the independent test set further illustrates the generalization capability and superiority of PScL-DDCFPred over existing predictors. In-depth analysis shows that the excellent performance of PScL-DDCFPred can be attributed to three critical factors, namely the effective combination of the DNN and DCF models, complementarity of global and local features, and use of the optimal feature sets selected by the integrative feature selection algorithm. Availability https://github.com/csbio-njust-edu/PScL-DDCFPred Supplementary information Supplementary data are available at Bioinformatics online.

PScL-2LSAESM: bioimage-based prediction of protein subcellular localization by integrating heterogeneous features with the two-level SAE-SM and mean ensemble method

“…In this section, to further illustrate the predictive power of PScL-2LSAESM, we performed experiments and compared its performance with that of the other existing protein subcellular localization predictors including PScL-DDCFPred ( Ullah et al , 2022 ), PScL-HDeep ( Ullah et al , 2021 ), SAE-RF ( Liu et al , 2020 ), SC-PSorter ( Shao et al , 2016 ) as well as the method proposed by Yang et al (2014) .…”

“…These included SLFs, LBP, CLBP, LETRIST and RICLBP with the dimensionalities of 840, 256, 906, 413 and 408, respectively. Previous studies have shown these features to be very effective in this field ( Ullah et al , 2021 , 2022 ; Xu et al , 2016 ). SLFs are the global features which includes 4-dimensional DNA distribution and 836-dimensional Haralick texture features ( Boland and Murphy, 2001 ; Ullah et al , 2022 ; Xu et al , 2013 ).…”

“…Previous studies have shown these features to be very effective in this field ( Ullah et al , 2021 , 2022 ; Xu et al , 2016 ). SLFs are the global features which includes 4-dimensional DNA distribution and 836-dimensional Haralick texture features ( Boland and Murphy, 2001 ; Ullah et al , 2022 ; Xu et al , 2013 ). SLFs are very useful for extracting the global texture information from images ( Boland and Murphy, 2001 ).…”

“…One useful way to utilize multiple feature sets is to simply concatenate all the feature sets in a simple serial fashion. Many protein subcellular localization prediction methods have utilized this simple serial integration strategy to integrate all the feature sets and subsequently develop a predictor ( Tahir et al , 2013 ; Ullah et al , 2022 ; Xu et al , 2016 , 2018 ). More recently, several studies have investigated new techniques other than simple serial integration ( Shao et al , 2016 ; Wang et al , 2022 ).…”

Pixel-level multimodal fusion deep networks for predicting subcellular organelle localization from label-free live-cell imaging

“…In comparison with conventional intelligence method, deep learning is able to perform a series of target recognitions, feature extraction, and analysis automatically, which makes it possible to automatically discover image–target features and explore feature levels and interaction ( Chen et al, 2016 ; Siu et al, 2020 ; Ullah et al, 2021 ; Li et al, 2022 ). The learning-enhanced cell optical image-analysis model is capable of acquiring the texture details from low-level source images and achieve higher resolution improvement for the label-free cell optical-imaging techniques ( Chen et al, 2016 ; Lee et al, 2020 ; Ullah et al, 2021 ; Ullah et al, 2022 ). The deep-learning pipeline of cell optical microscopy imaging can extract complex data representation in a hierarchical way, which is helpful to find hidden cell structures from the microscope images, such as the size of a single cell, the number of cells in a given area, the thickness of the cell wall, the spatial distribution between cells, and subcellular components and their densities ( Boslaugh and Watters, 2008 ; Donovan-Maiye et al, 2018 ; Falk et al, 2019 ; Manifold et al, 2019 ; Rezatofighi et al, 2019 ; Yao et al, 2019 ; Zhang et al, 2019 ; Lee et al, 2020 ; Voronin et al, 2020 ; Zhang et al, 2020 ; Chen et al, 2021a ; Gomariz et al, 2021 ; Manifold et al, 2021 ; Wang et al, 2022b ; Islam et al, 2022 ; Kim et al, 2022 ; Melanthota et al, 2022 ; Rahman et al, 2022 ; Ullah et al, 2022 ; Witmer and Bhanu, 2022 ).…”