Multi-Label Classification for Microbial Data
Overview
The service provides multi-label classification for microbial data, intended for analysis of samples that may contain multiple relevant sources or categories at the same time. Its main purpose is to support interpretation of complex microbial samples, especially mixed samples where a single-label assignment is not sufficient. The service enables automated identification of multiple relevant labels from one microbial sample and is particularly suitable for use cases where mixed microbial composition must be interpreted in a more detailed and realistic way. It is intended to support both research and applied workflows in microbial data analysis, especially where source attribution and proportion estimation are important. The service relies on a novel AI-based classification approach. It uses transformed microbial data representations prepared through prior preprocessing and feature extraction steps, and supports training, validation, and execution of hierarchical multi-label classification models. The workflow is designed to process both pure and mixed microbial data, account for multiple simultaneous labels, and provide outputs suitable for downstream interpretation and decision support. The implementation supports scalable execution, reproducible evaluation, and integration with the broader set of microbial analysis services available in SLAIF.
Outcome
Users obtain classification outputs in which microbial samples can be assigned multiple relevant labels, together with estimates that support interpretation of mixed microbial composition. The service enables more informative analysis of complex microbial data, supports identification of multiple contributing sources within a sample, and provides a basis for further analytical and decision-support workflows.
Service Card
| Category | AI products |
| Domain | Microbial Data Analysis, Bioinformatics |
| Vertical | Health and Biotechnology |
| Target users | Research Institutions, Health Institutions, Data Scientists, AI/ML Researchers |
| Service type | Service |
| Outputs | Multi-label classification results for microbial samples with proportion estimates supporting interpretation of mixed microbial composition. |
| Requirements | |
| Limitations | |
| Availability | Q3/2027 |
| Licensing | |
| Visual content | |
| Visual caption |