Managing biological variation data: modern approaches for study design and clinical application

Within- and between-subject biological variation estimates for the enumeration of lymphocyte deep immunophenotyping and monocyte subsets

OBJECTIVES

This study aimed to deliver biological variation (BV) estimates for 25 types of lymphocyte subpopulations subjected to deep immunophenotyping (memory T/B cells, regulatory T cells, etc.) and classical, intermediate, and nonclassical monocyte subsets based on the full spectrum flow cytometry (FS-FCM) and a Biological Variation Data Critical Appraisal Checklist (BIVAC) design.

METHODS

Samples were collected biweekly from 60 healthy Chinese adults over 10 consecutive two-week periods. Each sample was measured in duplicate within a single run for lymphocyte deep immunophenotyping and monocyte subset determination using FS-FCM, including the percentage (%) and absolute count (cells/μL). After trend adjustment, a Bayesian model was applied to deliver the within-subject BV (CVI) and between-subject BV (CVG) estimates with 95 % credibility intervals.

RESULTS

Enumeration (% and cells/μL) for 25 types of lymphocyte deep immunophenotyping and three types of monocyte subset percentages showed considerable variability in terms of CVI and CVG. CVI ranged from 4.23 to 47.47 %. Additionally, CVG ranged between 10.32 and 101.30 %, except for CD4+ effector memory T cells re-expressing CD45RA. No significant differences were found between males and females for CVI and CVG estimates. Nevertheless, the CVGs of PD-1+ T cells (%) may be higher in females than males. Based on the desired analytical performance specification, the maximum allowable imprecision immune parameter was the CD8+PD-1+ T cell (cells/μL), with 23.7 %.

CONCLUSIONS

This is the first study delivering BV estimates for 25 types of lymphocyte subpopulations subjected to deep immunophenotyping, along with classical, intermediate, and nonclassical monocyte subsets, using FS-FCM and adhering to the BIVAC design.

Quantitative evaluations of variations using the population mean as a baseline for bioinformatics interpretation

Objective

The purpose of this study were to establish a model of quantitative evaluation that uses the population mean as a baseline of variations and describe variations derived from different types and systems using new concepts.

Methods

The observed datasets, including measurement data and relative data, were transformed to 0-1.0 using the population mean. Datasets derived from different types (same category of dataset, different categories of datasets, and datasets with the same baseline) were transformed using different methods. The 'middle compared index' (MCI) was used to describe the change in magnitude as follows: [a/(a+b)+(1-b)/(2-a-b)-1]^1.7, where 'a' represents the number after the magnitude change and 'b' represents the number before the magnitude change. Actual data were used to observe the MCI's ability to evaluate variations quantitatively.

Results

When the value before the magnitude change was equal to that after the magnitude change, the MCI was equal to 0; when the value before the magnitude change was equal to 0 and that after the magnitude change was equal to 1, the MCI was equal to 1. This implies the MCI is valid. When the value before the magnitude change was 0 and that after the magnitude change was 0.5, or when the value before the magnitude change was 0.5 and that after the magnitude change was 1.0, each MCI was approximately equal to 0.5. The values derived from the absolute, ratio, and MCI methods were different, indicating that the MCI is an independent index.

Conclusion

The MCI perfectly performs as an evaluation model using the population mean as the baseline, and it may be more a reasonable index than the ratio or absolute methods. The MCI increases our understanding of quantitative variations in evaluation measures of association using new concepts.

Clinlabomics: leveraging clinical laboratory data by data mining strategies

The recent global focus on big data in medicine has been associated with the rise of artificial intelligence (AI) in diagnosis and decision-making following recent advances in computer technology. Up to now, AI has been applied to various aspects of medicine, including disease diagnosis, surveillance, treatment, predicting future risk, targeted interventions and understanding of the disease. There have been plenty of successful examples in medicine of using big data, such as radiology and pathology, ophthalmology cardiology and surgery. Combining medicine and AI has become a powerful tool to change health care, and even to change the nature of disease screening in clinical diagnosis. As all we know, clinical laboratories produce large amounts of testing data every day and the clinical laboratory data combined with AI may establish a new diagnosis and treatment has attracted wide attention. At present, a new concept of radiomics has been created for imaging data combined with AI, but a new definition of clinical laboratory data combined with AI has lacked so that many studies in this field cannot be accurately classified. Therefore, we propose a new concept of clinical laboratory omics (Clinlabomics) by combining clinical laboratory medicine and AI. Clinlabomics can use high-throughput methods to extract large amounts of feature data from blood, body fluids, secretions, excreta, and cast clinical laboratory test data. Then using the data statistics, machine learning, and other methods to read more undiscovered information. In this review, we have summarized the application of clinical laboratory data combined with AI in medical fields. Undeniable, the application of Clinlabomics is a method that can assist many fields of medicine but still requires further validation in a multi-center environment and laboratory.