Abstract:Proteins are a kind of biological macromolecules that play a vital role in building the cytoskeleton, being a catalyst in biological processes, and participating in the biological immune process. The loss and incorrect of the subcellular location of proteins prone to suffer from pathological disease. Therefore, predicting accurately subcellular location of proteins has become a hot research topic. In particular, proteins that occur in multiple subcellular locations have received little attention. Based on this, a novel protein multi-label subcellular location predictor called MFTr_Locator was developed to predict the subcellular location of multi-label proteins in multi-field immunohistochemistry (IHC) images. The proposed method consists of four main steps, namely, splitting the IHC images into multi-field patches, image feature extraction, feature selection, and multi-label classification. First, an IHC image was initially split into three scales using three-size slide windows, which represent the region with abundant protein expression in the current field size. In addition, subcellular location features (SLFs) and frequency domain operator were employed to quantify the multi-scaled IHC patch images, and stepwise discriminant analysis (SDA) was adopted to reduce feature dimension. Finally, a concatenated feature that included multi-field features from an IHC image was fed into the decoder of the Transformer to recognize subcellular locations of multi-label proteins. To validate the proposed method, a benchmark dataset comprising 2100 single-label and 1200 multi-label IHC images from the human protein atlas (HPA) was collected. The main contributions of our method are as follows: Firstly, we employ multi-field images instead of single-field images for subcellular location recognition. Secondly, we use Transformer's decoder as a multi-label classifier. From the experimental results, the subset accuracy of mixed labels reaches 79.7%. The promising results show that the developed model substantially outperforms current in-field methods and may provide protein subcellular location with model analysis tools effectively.

MFTr_Locator: a Novel Transformer Model Decoding Multi-Label Protein Subcellular Locations in Multi-Field ImmunohistoChemistry Image

Erratum to: Specific stressors in endonasal skull base surgery with and without navigation

Mmformer: Multimodal Medical Transformer for Incomplete Multimodal Learning of Brain Tumor Segmentation

Deep learning model for protein multi-label subcellular localization and function prediction based on multi-task collaborative training

Protein Subcellular Localization Prediction by Concatenation of Convolutional Blocks for Deep Features Extraction From Microscopic Images

Imbalanced multi-modal multi-label learning for subcellular localization prediction of human proteins with both single and multiple sites

Vislocas: Vision transformers for identifying protein subcellular mis-localization signatures of different cancer subtypes from immunohistochemistry images

Multilabel Learning for Protein Subcellular Location Prediction

MSlocPRED: deep transfer learning-based identification of multi-label mRNA subcellular localization

LncLocFormer: a Transformer-based deep learning model for multi-label lncRNA subcellular localization prediction by using localization-specific attention mechanism

SIFLoc: a self-supervised pre-training method for enhancing the recognition of protein subcellular localization in immunofluorescence microscopic images

Backbone 1H, 13C and 15N resonance assignments of the 39 kDa staphylococcal hemoglobin receptor IsdH

DMLDA-LocLIFT: Identification of multi-label protein subcellular localization using DMLDA dimensionality reduction and LIFT classifier

dMIL-Transformer: Multiple Instance Learning via Integrating Morphological and Spatial Information for Lymph Node Metastasis Classification

Extracting Cellular Location of Human Proteins Using Deep Learning

MFTrans: A Multi-Feature Transformer Network for Protein Secondary Structure Prediction

iDeepSubMito: identification of protein submitochondrial localization with deep learning

Hierarchical Transformer for Survival Prediction Using Multimodality Whole Slide Images and Genomics

Automated classification of protein subcellular localization in immunohistochemistry images to reveal biomarkers in colon cancer

iPCa-Former: A Multi-Task Transformer Framework for Perceiving Incidental Prostate Cancer

An Artificial Intelligence-Based Stacked Ensemble Approach for Prediction of Protein Subcellular Localization in Confocal Microscopy Images