Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Jun 1;34(11):1808-1816.
doi: 10.1093/bioinformatics/bty016.

QuantumClone: Clonal Assessment of Functional Mutations in Cancer Based on a Genotype-Aware Method for Clonal Reconstruction

Affiliations
Free PMC article

QuantumClone: Clonal Assessment of Functional Mutations in Cancer Based on a Genotype-Aware Method for Clonal Reconstruction

Paul Deveau et al. Bioinformatics. .
Free PMC article

Abstract

Motivation: In cancer, clonal evolution is assessed based on information coming from single nucleotide variants and copy number alterations. Nonetheless, existing methods often fail to accurately combine information from both sources to truthfully reconstruct clonal populations in a given tumor sample or in a set of tumor samples coming from the same patient. Moreover, previously published methods detect clones from a single set of variants. As a result, compromises have to be done between stringent variant filtering [reducing dispersion in variant allele frequency estimates (VAFs)] and using all biologically relevant variants.

Results: We present a framework for defining cancer clones using most reliable variants of high depth of coverage and assigning functional mutations to the detected clones. The key element of our framework is QuantumClone, a method for variant clustering into clones based on VAFs, genotypes of corresponding regions and information about tumor purity. We validated QuantumClone and our framework on simulated data. We then applied our framework to whole genome sequencing data for 19 neuroblastoma trios each including constitutional, diagnosis and relapse samples. We confirmed an enrichment of damaging variants within such pathways as MAPK (mitogen-activated protein kinases), neuritogenesis, epithelial-mesenchymal transition, cell survival and DNA repair. Most pathways had more damaging variants in the expanding clones compared to shrinking ones, which can be explained by the increased total number of variants between these two populations. Functional mutational rate varied for ancestral clones and clones shrinking or expanding upon treatment, suggesting changes in clone selection mechanisms at different time points of tumor evolution.

Availability and implementation: Source code and binaries of the QuantumClone R package are freely available for download at https://CRAN.R-project.org/package=QuantumClone.

Contact: gudrun.schleiermacher@curie.fr or valentina.boeva@inserm.fr.

Supplementary information: Supplementary data are available at Bioinformatics online.

Figures

Fig. 1.
Fig. 1.
Comparison of QuantumClone to existing methods. (A) NMI is used to assess the quality of variant clustering on simulated data, with a single parameter varying in each test. This measure evaluates correct assignment of two variants to the same cluster. (B)L2 average error is used to assess the error for each clustered variants between its simulated position and its reconstructed position. (C) Computational time necessary to complete the clustering with each algorithm. Default parameters: two tumor samples without contamination sequenced at 100×; 6 clones; 100 mutations used for clustering
Fig. 2.
Fig. 2.
Quality of clonal reconstruction for mutations located in regions of altered copy number. (A) NMI shows equivalent performances of pyClone and QuantumClone in diploid, triploid and tetraploid tumors, or nearly diploid (ND) tumors, whereas the average L2 error (B) shows significantly better performance of QuantumClone. (C) Parallel computing implemented in QuantumClone allows it to significantly decrease computational time and makes QuantumClone remarkably faster than pyClone
Fig. 3.
Fig. 3.
Assessment of the pipeline. (A) Overview of the general clonal reconstruction workflow: steps 1–3. (1) Variants are filtered to remove false positive calls; stringent filters are used to produce mutations that are further employed for clonal reconstruction (step 2), tolerant filters are used to detect functional mutations. (2) Variants that pass stringent filters and have genotype information assigned to the corresponding genomic loci are used as input to QuantumClone to reconstruct clonal populations. (3) Finally, possibly damaging mutations belonging to frequently altered pathways are mapped to the reconstructed clones. Quality of reconstruction. The pipeline aforementioned (two step), or a clustering using all variants called (classic) or a pipeline using only variants of biological interest and variants of high quality (selective) are assessed in terms of NMI (B), average L2 error (C) or computational time (D). The pipelines are evaluated on 20 simulations (Section 2)
Fig. 4.
Fig. 4.
Annotation of clones in neuroblastoma and pathway enrichment analysis. (A) Illustration with data from patient NB1361 of the rules for assignment of variants to (i) the ancestral clone (cellular prevalence of the mutation cluster exceeds 70% both at diagnosis and relapse), (ii) clones expanding after the treatment (cellular prevalence of the mutation cluster increases at least two-fold at relapse) and (iii) shrinking clones (cellular prevalence of such mutation clusters decreases at least two-fold). Here, evaluated cellular prevalence values higher than 1 were set to 1. (B) Evolution of the total number of functional variants for enriched maps and modules, across all 19 patients. The majority of modules show an increase in the number of functional variants between the two time points
Fig. 5.
Fig. 5.
Ancestral, shrinking and expanding clones exhibit different mutation patterns in neuroblastoma relapse tumors. (A) Functional mutation rate is higher in shrinking and expanding clones compared to the ancestral ones. We define the functional mutation rate as a ratio of the number of functional mutations to the number of high fidelity variants. For a given gene module the number of functional mutations in each patient is supposed to linearly depend on the product of the module size and the total number of detected variants. Therefore, we used the product of the module size and number of high fidelity variants as a covariate in a linear regression model evaluating functional mutation rate for neuroblastoma tumors. The rate was defined as the slope of the linear regression. (B) Given the differences in functional mutation rates observed in neuroblastoma relapse tumors we propose the following model for clonal selection in this type of cancer: (1) Clones with high functional mutation rate (red) disappear after the chemotherapy; lower mutational burden provides an advantage in escape from treatment; (2) lower values for functional mutation rate in clones expanding at relapse (blue) compared to the shrinking clones (red) is due to a lower frequency of functional mutations before treatment, followed by a gradual accumulation of functional mutations at relapse. From top to bottom: the number of variants in the clone, number of functional variants in the clone, and population size in the tumor

Similar articles

See all similar articles

Cited by 3 articles

References

    1. Adzhubei I. et al. (2013) Predicting functional effect of human missense mutations using PolyPhen-2. Curr. Protocol. Hum. Genet., 0 7, Unit7.20. - PMC - PubMed
    1. Chen G. et al. (2015) Targeting the adaptability of heterogeneous aneuploids. Cell, 160, 771–784. - PMC - PubMed
    1. Deshwar A.G. et al. (2015) PhyloWGS: reconstructing subclonal composition and evolution from whole-genome sequencing of tumors. Genome Biol., 16, 35.. - PMC - PubMed
    1. Deveau P. et al. (2016) Calculating biological module enrichment or depletion and visualizing data on large-scale molecular maps with ACSNMineR and RNaviCell packages. R J., 8, 293–306.
    1. Eleveld T.F. et al. (2015) Relapsed neuroblastomas show frequent RAS-MAPK pathway mutations. Nat. Genet., 47, 864–871. - PMC - PubMed

Publication types

Feedback