Helmholtz Zentrum München and the Technische Universität München have recently collaborated to develop a comprehensive system that consolidates all of the unicellular or single-cells in one place. The purpose of the new single-cell atlas was to provide all of the information the experts need about their desired cell in one place. The atlas uses machine learning and AI for the best integration of single-cell queries in just one architecture.
Single-Cell Architecture Surgery (scArche)
The new cell atlas is called 'scArche,' a shorter name for single-cell architecture surgery. scArche uses a unique algorithm based on transfer learning for more effective processing of single-cell queries without taking too much time and effort on harvesting information or sharing raw data. According to a PhysOrg report, the new algorithm has been tested for numerous studies. The capacity of the scArche single-cell reference model was also utilized to examine other conditions, including COVID-19.
Munich's biology expert and author of the study Mohammad Lotfollahi said that the single-cell reference algorithm is advantageous compared to the traditional raw data sharing between laboratories and institutions. The single-cell atlas can provide genomics information that saves up the query history without breaching any data, proving security for anonymity. According to Lotfollahi, the single-cell atlas can quickly provide new data sets, making them easier to use when interpreting and gathering new datasets.
The single-cell reference development was a big step for painstaking studies that require an excessive amount of work and examinations. The authors wrote on the published research that smaller-scale studies on genomics are now made easier through the help of an AI-powered large-scale single-cell atlas that will effectively provide detailed analysis and reference to any studies.
scArche: Limitations, Solutions, and COVID-19
The single-cell atlas development, although perfectly working, still have minor limitations. The researchers wrote that the reference data is not yet fully consolidated because of the limited resource, data restrictions, and differences in the datasets between various institutions and laboratories that could have distinct methodologies and interpretations based on their own protocol or principle.
According to the experts, data integration is utilized to close the gap between the varying information over single-cells. However, to match every batch of references, the single-cell atlas would require a number of relevant data from other sources that require privilege or access. This phase would need a complex legal process for the experts to get valuable single-cell information.
In case that other single-cell data is proven to be unattainable, examinations can take place, but the analysis and integration of new data are expected to run for a long time and may be exhaustive.
The scArche was used for COVID-19 studies that involve a bronchial examination. The experts were able to compare the lung cells of an infected subject to a healthy subject through single-cell transcriptomics. Through the use of the scArches algorithm, the scientists could define the diseased cells from the references for COVID-19. The whole coverage of the AI-powered single-cell atlas was published in the journal Nature Biotechnology, titled "Mapping single-cell data to reference atlases by transfer learning."
Check out more news and information on Biology in Science Times.