Editorial: The Human Cell Atlas. What Is It and Where Could It Take Us?

An estimated 37 trillion cells comprise the human body, each with its specialized role [1]. The Human Cell Atlas (HCA) Consortium was founded in 2016 as an open global initiative to map each cell type in the human body and create a three-dimensional (3-D) atlas [2]. As of December 2024, 18 Biological Networks are assembling the first draft of the HCA from organs, tissues, and organ systems, including the heart, lung, liver, and immune system [2]. Although the completed first version of the HCA should be released within a year, possibly two, the HCA Biological Networks are making the atlases available on the HCA Data Portal as they are released [3]. The freely available data and cell maps will help transform future healthcare by improving the understanding of tissue-specific human cell biology in health and disease [3]. Since 2016, the Consortium has grown to include more than 3,200 members from more than 1,700 institutes and now involves 99 countries to allow data from diverse geographic and ethnic groups and age ranges [3]. The HCA has started to collect cells, study the biological characteristics of each collected cell, determine the level in various organs of the human body, and estimate the extent of variation across individuals [4]. As of December 2024, 9,200 donors, 62.7 million cells, and 478 projects have been included in the HCA, which has currently mapped 18 biological networks (Table 1) [3].

Reference maps with sufficient variation are a first step towards inferring the possible causes of variation across ancestry, demographic variables, and environmental exposures, understanding biological mechanisms, and developing disease models and treatments more generalizable to whole human populations [4]. An important ethical consideration is that written informed consent is obtained from each volunteer willing to participate in HCA after the volunteer understands the risks and benefits of participation and before collecting samples [4]. Adequate safeguards are in place to minimize risks and discomfort associated with sample collection [4]. The coding process removes all direct identifiers, such as name and address, to maintain donor anonymity [4].

The data collected from each participant includes age at the time of collection, gender, information on how the cells were collected, and the primary data generated in the laboratory, such as levels of expression of genes in single cells [4]. Depending on the study, general information about the health of the study participants health and environment is included [4]. The HCA has created a cloud-based Data Portal that stores data contributed by HCA collaborators with a Data Explorer interface to help researchers find and explore the HCA datasets [3]. Anonymized data from the HCA is coded to mark individual participants and placed on a globally accessible, public domain database for advancing science by HCA investigators and anyone interested in doing so by analyzing the data [4]. Therefore, the HCA follows the ethos of other large-scale resources that share and disseminate biomedical datasets, including the Human Genome Project and the 1000 Genomes Project and local or regional initiatives, such as the UK Biobank [4].

The methods used by research teams as part of the HCA Consortium have been aimed at the single-cell level [5]. Single-cell transcriptomics, spatial genomics, computational and artificial intelligence, and machine learning (AI/MCL) techniques have been used to identify active genes and cellular mechanisms [6]. Advances in molecular profiling methods and new computational methods have allowed the construction of the HCA, a progression from data collection to atlas integration [6]. Single-cell technologies have been the basis for developing human cell atlases, including the HCA, and have allowed scientists to interrogate genes and cellular features cell by cell rather than taking the average features from many cells that comprise organs and tissues [6]. These single-cell technologies have allowed the HCA to profile hundreds of millions of single cells, resulting in up to 450 published studies in 2024 [5]. Research studies have reached a critical mass, resulting in recent publications on cell mapping, data integration, and predictive modeling [5,7,8]. For example, Ergen and colleagues have recently developed the popV tool incorporating eight automated cell-annotation tools into one platform, which can be expanded as more tools become available [5]. Recently generated single-cell RNA sequencing data can be loaded into the popV platform, which will then apply consensus (a vote) for the most likely cell identity and combine it with a certainty score [5]. There are plans to incorporate popV into the HCA Cell Annotation Platform user interface. To facilitate cell identity, Heimberg and colleagues have developed SCimilarity software to take a cellular profile of interest and identify related cells [7]. AI/MCL can overcome the cost of high-resolution or high-throughput single-cell studies to extrapolate single-cell data by using scSemiProfiler, which uses bulk data and generative AI to produce the likely spread of single-cell profiles [8].

An anticipated development will be to use AI/MCL to generate spatially resolved single-cell RNA sequencing data from routine archival diagnostic histopathology tissue sections stained with hematoxylin and eosin (H&E) [9,10]. The cell and tissue morphology and their associations, seen with histochemical staining patterns, are likely to be associated with molecular features, such as gene expression [9,10]. For this reason, the Single-Cell omics from Histology Analysis Framework (SCHAF) program is being developed at the Massachusetts Institute of Technology stained tissue sections and single-cell RNA profiles from an adjacent tissue section [10]. A further recent development is the use of computer models to incorporate multiple data from the same sample, as in multiDGD, which models biology using RNA expression and chromatin-accessibility data from the same cells [11]. Future developments include using single-cell technologies to screen therapeutic drugs and identify small molecules that could replace more expensive biological therapies.

There are several ways in which cell atlases, including the HCA, have begun to provide important biological insights, which are likely to lead to future medical benefits [6,12]. A critical stage has been reached in the progress of the HCA, where advances in spatial and molecular profiling methods and new computational methods are promoting the use of AI/MCL [6]. Cell atlases, including the HCA, have begun to reveal valuable biological insights and are anticipated to soon lead to significant medical benefits [6]. Cell atlases have become established as a consensus of cells, as 3-D maps of cells connecting genotypic causes to phenotype effects of disease, and as 4-D maps of cell and tissue development [6]. The HCA represents a foundation model of biology that has a vital role in the journey to unify all these aspects to transform medicine.

The HCA has the potential to benefit humanity by providing a path to equitable research, involvement of previously underrepresented populations, sharing research benefits, and reducing research bias [4]. A key objective of the HCA is to investigate cellular variation in healthy tissues to understand health and disease [4]. When genetic, environmental, and experiential factors vary, collecting cells from various organs from diverse cohorts of people is crucial [4]. Also, measuring diversity and variation within healthy tissues allows a better understanding of the range of cells across humans [4]. The HCA has major implications for future biomedical research and its commitment to humanity through scientific engagement and regional networks [4]. The HCA has created the Equity Working Group (EqWG) to develop outreach, training, and education [4]. The HCA Ethics Working Group has supported efforts to implement the HCA globally via an ethics toolkit, which provides researchers with templates (consent form templates and information pamphlets) and governance documents [13,14].

Conclusions

The identification and validation of further research tools and analytics with single-cell applications will likely emerge from the HCA project researchers and collaborators. The reference maps created by the HCA will help researchers understand whether each person afflicted with a disease, such as lung cancer, has common biological or cellular changes in the affected organ. Knowledge gained from this work will result in a better understanding of health and the management and care of individual patients. To achieve this ultimate goal, the reference maps must reflect biological and cellular variation in various organs among healthy individuals from diverse backgrounds, including variable genetic ancestries, gender, age, geography, environment, and lifestyle. The encyclopedia of cells that the HCA is creating will be an important step towards providing better care to individual patients, which will benefit all of humanity.