Author Correction: Analysis of the Human Protein Atlas Image Classification competition.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Publisher Correction: Analysis of the Human Protein Atlas Image Classification competition.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Publisher Correction: Analysis of the Human Protein Atlas Image Classification competition.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Analysis of the Human Protein Atlas Image Classification competition.

Pinpointing subcellular protein localizations from microscopy images is easy to the trained eye, but challenging to automate. Based on the Human Protein Atlas image collection, we held a competition to identify deep learning solutions to solve this task. Challenges included training on highly imbalanced classes and predicting multiple labels per image. Over 3 months, 2,172 teams participated. Despite convergence on popular networks and training techniques, there was considerable variety among the solutions. Participants applied strategies for modifying neural networks and loss functions, augmenting data and using pretrained networks. The winning models far outperformed our previous effort at multi-label classification of protein localization patterns by ~20%. These models can be used as classifiers to annotate new images, feature extractors to measure pattern similarity or pretrained networks for a wide range of biological applications.

Deep learning is combined with massive-scale citizen science to improve large-scale image classification.

Pattern recognition and classification of images are key challenges throughout the life sciences. We combined two approaches for large-scale classification of fluorescence microscopy images. First, using the publicly available data set from the Cell Atlas of the Human Protein Atlas (HPA), we integrated an image-classification task into a mainstream video game (EVE Online) as a mini-game, named Project Discovery. Participation by 322,006 gamers over 1 year provided nearly 33 million classifications of subcellular localization patterns, including patterns that were not previously annotated by the HPA. Second, we used deep learning to build an automated Localization Cellular Annotation Tool (Loc-CAT). This tool classifies proteins into 29 subcellular localization patterns and can deal efficiently with multi-localization proteins, performing robustly across different cell types. Combining the annotations of gamers and deep learning, we applied transfer learning to create a boosted learner that can characterize subcellular protein distribution with F1 score of 0.72. We found that engaging players of commercial computer games provided data that augmented deep learning and enabled scalable and readily improved image classification.

A subcellular map of the human proteome.

Resolving the spatial distribution of the human proteome at a subcellular level can greatly increase our understanding of human biology and disease. Here we present a comprehensive image-based map of subcellular protein distribution, the Cell Atlas, built by integrating transcriptomics and antibody-based immunofluorescence microscopy with validation by mass spectrometry. Mapping the in situ localization of 12,003 human proteins at a single-cell level to 30 subcellular structures enabled the definition of the proteomes of 13 major organelles. Exploration of the proteomes revealed single-cell variations in abundance or spatial distribution and localization of about half of the proteins to multiple compartments. This subcellular map can be used to refine existing protein-protein interaction networks and provides an important resource to deconvolute the highly complex architecture of the human cell.

Antibody Validation in Bioimaging Applications Based on Endogenous Expression of Tagged Proteins.

Antibodies are indispensible research tools, yet the scientific community has not adopted standardized procedures to validate their specificity. Here we present a strategy to systematically validate antibodies for immunofluorescence (IF) applications using gene tagging. We have assessed the on- and off-target binding capabilities of 197 antibodies using 108 cell lines expressing EGFP-tagged target proteins at endogenous levels. Furthermore, we assessed batch-to-batch effects for 35 target proteins, showing that both the on- and off-target binding patterns vary significantly between antibody batches and that the proposed strategy serves as a reliable procedure for ensuring reproducibility upon production of new antibody batches. In summary, we present a systematic scheme for antibody validation in IF applications using endogenous expression of tagged proteins. This is an important step toward a reproducible approach for context- and application-specific antibody validation and improved reliability of antibody-based experiments and research data.

RNA deep sequencing as a tool for selection of cell lines for systematic subcellular localization of all human proteins.

One of the major challenges of a chromosome-centric proteome project is to explore in a systematic manner the potential proteins identified from the chromosomal genome sequence, but not yet characterized on a protein level. Here, we describe the use of RNA deep sequencing to screen human cell lines for RNA profiles and to use this information to select cell lines suitable for characterization of the corresponding gene product. In this manner, the subcellular localization of proteins can be analyzed systematically using antibody-based confocal microscopy. We demonstrate the usefulness of selecting cell lines with high expression levels of RNA transcripts to increase the likelihood of high quality immunofluorescence staining and subsequent successful subcellular localization of the corresponding protein. The results show a path to combine transcriptomics with affinity proteomics to characterize the proteins in a gene- or chromosome-centric manner.

Systematic validation of antibody binding and protein subcellular localization using siRNA and confocal microscopy.

We have developed a platform for validation of antibody binding and protein subcellular localization data obtained from immunofluorescence using siRNA technology combined with automated confocal microscopy and image analysis. By combining the siRNA technology with automated sample preparation, automated imaging and quantitative image analysis, a high-throughput assay has been set-up to enable confirmation of accurate protein binding and localization in a systematic manner. Here, we describe the analysis and validation of the subcellular location of 65 human proteins, targeted by 75 antibodies and silenced by 130 siRNAs. A large fraction of (80%) the subcellular locations, including locations of several previously uncharacterized proteins, could be confirmed by the significant down-regulation of the antibody signal after the siRNA silencing. A quantitative analysis was set-up using automated image analysis to facilitate studies of targets found in more than one compartment. The results obtained using the platform demonstrate that siRNA silencing in combination with quantitative image analysis of antibody signals in different compartments of the cells is an attractive approach for ensuring accurate protein localization as well as antibody binding using immunofluorescence. With a large fraction of the human proteome still unexplored, we suggest this approach to be of great importance under the continued work of mapping the human proteome on a subcellular level.

Mapping the subcellular protein distribution in three human cell lines.

The subcellular locations of proteins are closely related to their function and constitute an essential aspect for understanding the complex machinery of living cells. A systematic effort has been initiated to map the protein distribution in three functionally different cell lines with the aim to provide a subcellular localization index for at least one representative protein from all human protein-encoding genes. Here, we present the results of more than 3500 proteins mapped to 16 subcellular compartments. The results indicate a ubiquitous protein expression with a majority of the proteins found in all three cell lines and a large portion localized to two or more compartments. The inter-relationships between the subcellular compartments are visualized in a protein-compartment network based on all detected proteins. Hierarchical clustering was performed to determine how closely related the organelles are in terms of protein constituents and compare the proteins detected in each cell type. Our results show distinct organelle proteomes, well conserved across the cell types, and demonstrate that biochemically similar organelles are grouped together.