# Limitations and optimizations of cellular lineages tracking

## TL;DR

This paper explores how to optimize tracking of cell lineages using genetic barcodes while balancing accuracy and the number of traceable lineages.

## Contribution

The study introduces a mathematical model and simulations to determine optimal parameters for lineage tracking under resource constraints.

## Key findings

- Increasing barcode insertion probability can reduce lineage inference accuracy due to reading errors.
- There is a trade-off between the number of traceable lineages and the accuracy of lineage identification.
- Optimal experimental parameters depend on population size and barcode pool complexity.

## Abstract

Tracking cellular lineages using genetic barcodes provides insights across biology and has become an important tool. However, barcoding strategies remain ad hoc. We show that elevating barcode insertion probability and thus increasing the average number of barcodes within the cells, adds to the number of traceable lineages but may decrease the accuracy of lineages inference due to reading errors. We establish the trade-off between accuracy in tracing lineages and the total number of traceable lineages, and find optimal experimental parameters under limited resources concerning the populations size of tracked cells and barcode pool complexity.

Many biological aspects can be examined using individual cellular lineages. For example, it allows us to investigate stem cell differentiation, cellular cooperation, stability of a phenotype, and more. To do so, the cells of interest are tagged with heritable identifiers called barcodes. One of the most common methods to label and track numerous lineages uses stochastic and combinatorial tagging. Here we investigate some properties of this random barcode labeling using a simple model, its mathematical analysis, and simulation. In particular, we examine the number of traceable lineages and the accuracy of lineages identification, while varying the initial barcode pool size, the labeling probability, and the barcode reading errors. We show a possible tradeoff between the accuracy of lineage identification and the number of tagged cells. Accordingly, careful planning of an experiment - corresponding to the required accuracy and needed number of tracked lineages - will be informed by our approach.