Algorithmic approaches to clonal reconstruction in heterogeneous cell populations

Ismail, Wazim Mohammed; Nzabarushimana, Etienne; Tang, Haixu

doi:10.1007/s40484-019-0188-3

Cited by 8 publications

(4 citation statements)

References 84 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Algorithmic approaches to characterizing the clonal evolutionary history of an evolving cell population is critical not only for understanding the mechanisms of evolution in microbes and microbial communities ( [2,9,10,17,20,25]), but also for addressing important problems in cancer genomics, e.g., for reconstructing tumor genomes and predicting putative driver mutations ( [18,21]). The algorithmic problem of clonal construction demonstrates several combinatorial and probabilistic nature of the problem [22,13,6,14], indicating some useful avenues to tackle the computational challenges presented. In this paper, we present ClonalTREE2, a scalable algorithm for clonal reconstruction from time course genomic sequencing data under a maximum likelihood framework.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

A scalable algorithm for clonal reconstruction from sparse time course genomic sequencing data

Ismail

Tang²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Long-term evolution experiments (LTEEs) reveal the dynamics of clonal compositions in an evolving bacterial population over time. Accurately inferring the haplotypes - the set of mutations that identify each clone, as well as the clonal frequencies and evolutionary history in a bacterial population is useful for the characterization of the evolutionary pressure on multiple correlated mutations instead of that on individual mutations. Here, we study the computational problem of reconstructing the haplotypes of bacterial clones from the variant allele frequencies (VAFs) observed during a time course in a LTEE. Previously, we formulated the problem using a maximum likelihood approach under the assumption that mutations occur spontaneously, and thus the likelihood of a mutation occurring in a specific clone is proportional to the frequency of the clone in the population when the mutation occurs. We also developed several heuristic greedy algorithms to solve the problem, which were shown to report accurate results of clonal reconstruction on simulated and real time course genomic sequencing data in LTEE. However, these algorithms are too slow to handle sparse time course data when the number of novel mutations occurring during the time course are much greater than the number of time points sampled. In this paper, we present a novel scalable algorithm for clonal reconstruction from sparse time course data. We employed a statistical method to estimate the sampling variance of VAFs derived from low coverage sequencing data and incorporated it into the maximum likelihood framework for clonal reconstruction on noisy sequencing data. We implemented the algorithm (named ClonalTREE2) and tested it using simulated and real sparse time course genomic sequencing data. The results showed that the algorithm was fast and achieved near-optimal accuracy under the maximum likelihood framework for the time course data involving hundreds of novel mutations at each time point. The source code of ClonalTREE2 is available at https://github.com/COL-IU/ClonalTREE2.

show abstract

Section: Discussionmentioning

confidence: 99%

“…It is non-trivial to reconstruct the clonal structure from the variant allele frequencies (VAF) [14]. Previously, we formulated the time course clonal reconstruction problem using a maximum likelihood framework [15], and proposed several heuristic greedy algorithms to solve the problem.…”

Section: Introductionmentioning

confidence: 99%

A scalable algorithm for clonal reconstruction from sparse time course genomic sequencing data

Ismail

Tang²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…High-throughput biological technologies, such as microarrays, RNA-seq and whole-genome bisulfite sequencing, provide us informative approaches to investigate various biological samples collected from laboratories or clinical trials [1,2]. However, plenty of these biological samples are complex mixtures of many different cell types without knowing their accurate proportions.…”

Section: Introductionmentioning

confidence: 99%

ARIC: accurate and robust inference of cell type proportions from bulk gene expression or DNA methylation data

Zhang¹,

Xu²,

Qiao³

et al. 2021

Briefings in Bioinformatics

View full text Add to dashboard Cite

Quantifying cell proportions, especially for rare cell types in some scenarios, is of great value in tracking signals associated with certain phenotypes or diseases. Although some methods have been proposed to infer cell proportions from multicomponent bulk data, they are substantially less effective for estimating the proportions of rare cell types which are highly sensitive to feature outliers and collinearity. Here we proposed a new deconvolution algorithm named ARIC to estimate cell type proportions from gene expression or DNA methylation data. ARIC employs a novel two-step marker selection strategy, including collinear feature elimination based on the component-wise condition number and adaptive removal of outlier markers. This strategy can systematically obtain effective markers for weighted $\upsilon$-support vector regression to ensure a robust and precise rare proportion prediction. We showed that ARIC can accurately estimate fractions in both DNA methylation and gene expression data from different experiments. We further applied ARIC to the survival prediction of ovarian cancer and the condition monitoring of chronic kidney disease, and the results demonstrate the high accuracy and robustness as well as clinical potentials of ARIC. Taken together, ARIC is a promising tool to solve the deconvolution problem of bulk data where rare components are of vital importance.

show abstract

“…Therefore, it is difficult to distinguish true passenger mutations from noise with high certainty, whereas these are of importance especially when inferring subclones that have not necessarily each gained novel driver mutations. Finally, the lack of ground truth data makes it difficult to validate the true performance of deconvolution-based methods [86].…”

Section: Deconvolution Methodsmentioning

confidence: 99%

Learning to better understand:

Nieboer¹

View full text Add to dashboard Cite

Introduction 2 TargetClone: a multi-sample approach for reconstructing subclonal evolution of tumors 3 svMIL: Predicting the pathogenic effect of TAD boundary-disrupting somatic structural variants through multiple instance learning 4 Predicting pathogenic non-coding SVs disrupting the 3D genome in 1,646 whole cancer genomes using multiple instance learning 5 Pan-cancer gene deficiency status prediction in 4,069 whole cancer genomes using machine learning 6 Discussion

show abstract

Algorithmic approaches to clonal reconstruction in heterogeneous cell populations

Cited by 8 publications

References 84 publications

A scalable algorithm for clonal reconstruction from sparse time course genomic sequencing data

A scalable algorithm for clonal reconstruction from sparse time course genomic sequencing data

ARIC: accurate and robust inference of cell type proportions from bulk gene expression or DNA methylation data

Learning to better understand:

Contact Info

Product

Resources

About