Description
Nuclear chromatin patterns have been used historically to distinguish between different developmental stages and lineages of white blood cells. While it is common to characterize cells based on arbitrary ‘open’ or ‘closed’ chromatin patterns, quantification of chromatin data is lacking. By extracting nuclei from a database of white blood cells and running fractal analyses using TWOMBLI, we were able to produce meaningful data quantifying chromatin patterns. Our data were put through a random forest algorithm which grouped each point based on relationship probability. The algorithm compared immature and mature cell types as well as cells of similar maturity and differing lineage (i.e. mature monocytes, lymphocytes, neutrophils, etc.). The data were organized into tables in order to visualize the relationship between chromatin pattern, stage of maturity, and cell lineage. The random forest algorithm was able to distinguish immature cells (blasts) from mature cells with a success rate of 93.9%. It was able to distinguish mature cells from immature cells with a success rate of 96.9%. Promyelocytes were found to be the most unique among those in the blast category as they were correctly identified 99.1% of the time. Monocytes and neutrophils were the most unique in the mature category, being correctly identified 90.9% and 92.0% of the time, respectively. Most patterns were not accurately categorized when comparing cells of different lineages at equivalent stages of maturity. Cells from myeloid and lymphoid lineages did not show chromatin patterns unique enough to accurately distinguish between them. The data have shown that chromatin patterns alone can be used to distinguish between immature and mature white blood cells with a high level of accuracy. Comparing cells of similar maturity and differing lineage shows that chromatin patterns alone are not sufficient to accurately differentiate based on cell type. Cells from myeloid and lymphoid lineages did not show chromatin patterns unique enough to accurately distinguish between them. This suggests that more research needs be performed to determine other quantifiable factors that can be used to differentiate cells based on lineage.
Disciplines
Clinical Trials | Health and Medical Administration | Health and Physical Education | Health Services Research | Hematology | Laboratory Medicine | Medical Education | Medical Immunology | Medical Sciences | Medicine and Health Sciences | Nursing | Primary Care | Scholarship of Teaching and Learning | Teacher Education and Professional Development
Document Type
Event
Recommended Citation
Young, Paul; Cordner, Ryan; and Gordhamer, Abigail, "Chromatin Pattern Quantification for the Differentiation of White Blood Cells" (2024). Annual Research Symposium. 16.
https://ecommons.roseman.edu/researchsymposium/2024/basic_sciences/16
Included in
Clinical Trials Commons, Health and Medical Administration Commons, Health and Physical Education Commons, Health Services Research Commons, Hematology Commons, Laboratory Medicine Commons, Medical Education Commons, Medical Immunology Commons, Nursing Commons, Primary Care Commons, Scholarship of Teaching and Learning Commons, Teacher Education and Professional Development Commons
Chromatin Pattern Quantification for the Differentiation of White Blood Cells
Nuclear chromatin patterns have been used historically to distinguish between different developmental stages and lineages of white blood cells. While it is common to characterize cells based on arbitrary ‘open’ or ‘closed’ chromatin patterns, quantification of chromatin data is lacking. By extracting nuclei from a database of white blood cells and running fractal analyses using TWOMBLI, we were able to produce meaningful data quantifying chromatin patterns. Our data were put through a random forest algorithm which grouped each point based on relationship probability. The algorithm compared immature and mature cell types as well as cells of similar maturity and differing lineage (i.e. mature monocytes, lymphocytes, neutrophils, etc.). The data were organized into tables in order to visualize the relationship between chromatin pattern, stage of maturity, and cell lineage. The random forest algorithm was able to distinguish immature cells (blasts) from mature cells with a success rate of 93.9%. It was able to distinguish mature cells from immature cells with a success rate of 96.9%. Promyelocytes were found to be the most unique among those in the blast category as they were correctly identified 99.1% of the time. Monocytes and neutrophils were the most unique in the mature category, being correctly identified 90.9% and 92.0% of the time, respectively. Most patterns were not accurately categorized when comparing cells of different lineages at equivalent stages of maturity. Cells from myeloid and lymphoid lineages did not show chromatin patterns unique enough to accurately distinguish between them. The data have shown that chromatin patterns alone can be used to distinguish between immature and mature white blood cells with a high level of accuracy. Comparing cells of similar maturity and differing lineage shows that chromatin patterns alone are not sufficient to accurately differentiate based on cell type. Cells from myeloid and lymphoid lineages did not show chromatin patterns unique enough to accurately distinguish between them. This suggests that more research needs be performed to determine other quantifiable factors that can be used to differentiate cells based on lineage.