Examining Tumor Phylogeny Inference in Noisy Sequencing Data Kiran - PowerPoint PPT Presentation

Examining Tumor Phylogeny Inference in Noisy Sequencing Data Kiran Tomlinson and Layla Oesper Department of Computer Science, Carleton College Dec. 4, 2018 Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 1 / 24

Clonal theory (Nowell 1976) Mutation Time Tumor population Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 2 / 24

Clonal theory (Nowell 1976) Mutation Mutations Cell Time Heterogeneous tumor Tumor population Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 2 / 24

Clonal theory (Nowell 1976) Mutation Mutations Cell Time Clonal tree Heterogeneous tumor Tumor population Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 2 / 24

Inferring tumor phylogeny How can we reconstruct a tumor’s clonal tree from its genome? ? Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 3 / 24

Inferring tumor phylogeny How can we reconstruct a tumor’s clonal tree from its genome? ? Why is this important? Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 3 / 24

Inferring tumor phylogeny How can we reconstruct a tumor’s clonal tree from its genome? ? Why is this important? 1 Personalized medicine (Greaves 2015), (McGranahan and Swanton 2017) Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 3 / 24

Inferring tumor phylogeny How can we reconstruct a tumor’s clonal tree from its genome? ? Why is this important? 1 Personalized medicine (Greaves 2015), (McGranahan and Swanton 2017) 2 Improved understanding of cancer development Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 3 / 24

Outline Background 1 Previous work Bulk sequencing data ISA AncesTree Methods 2 Results 3 Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 4 / 24

Previous work …GAT… …GTT… Single nucleotide variants (SNV) only: PhyloSub (Jiao et al. 2014) Rec-BTP (Hajirasouliha et al. 2014) AncesTree (El-Kebir et al. 2015) CITUP (Malikic et al. 2015) LICHeE (Popic et al. 2015) BitPhylogeny (Yuan et al. 2015) Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 5 / 24

Previous work …GAT… …GTT… Single nucleotide variants (SNV) SNVs and CNAs/structural variants: only: SubcloneSeeker (Qiao et al. 2014) PhyloSub (Jiao et al. 2014) PhyloWGS (Deshwar et al. 2015) Rec-BTP (Hajirasouliha et al. 2014) SPRUCE (El-Kebir et al. 2016) AncesTree (El-Kebir et al. 2015) Canopy (Jiang et al. 2016) CITUP (Malikic et al. 2015) PASTRI (Satas and Raphael 2017) LICHeE (Popic et al. 2015) BitPhylogeny (Yuan et al. 2015) Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 5 / 24

Previous work …GAT… …GTT… Single nucleotide variants (SNV) SNVs and CNAs/structural variants: only: SubcloneSeeker (Qiao et al. 2014) PhyloSub (Jiao et al. 2014) PhyloWGS (Deshwar et al. 2015) Rec-BTP (Hajirasouliha et al. 2014) SPRUCE (El-Kebir et al. 2016) AncesTree (El-Kebir et al. 2015) Canopy (Jiang et al. 2016) CITUP (Malikic et al. 2015) PASTRI (Satas and Raphael 2017) LICHeE (Popic et al. 2015) Single-cell and bulk data: BitPhylogeny (Yuan et al. 2015) Single-cell sequencing data: ddClone (Salehi et al. 2017) B-SCITE (Malikic et al. 2018) OncoNEM (Ross et al. 2016) SCITE (Jahn et al. 2016) and many more.... SiFit (Zafar et al. 2017) Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 5 / 24

Previous work …GAT… …GTT… Single nucleotide variants (SNV) SNVs and CNAs/structural variants: only : SubcloneSeeker (Qiao et al. 2014) PhyloSub (Jiao et al. 2014) PhyloWGS (Deshwar et al. 2015) Rec-BTP (Hajirasouliha et al. 2014) SPRUCE (El-Kebir et al. 2016) AncesTree (El-Kebir et al. 2015) Canopy (Jiang et al. 2016) CITUP (Malikic et al. 2015) PASTRI (Satas and Raphael 2017) LICHeE (Popic et al. 2015) Single-cell and bulk data: BitPhylogeny (Yuan et al. 2015) Single-cell sequencing data: ddClone (Salehi et al. 2017) B-SCITE (Malikic et al. 2018) OncoNEM (Ross et al. 2016) SCITE (Jahn et al. 2016) and many more.... SiFit (Zafar et al. 2017) Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 5 / 24

Bulk sequencing data Aligned reads Sample 2 Sample 1 (S2) (S1) S1 S2 Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 6 / 24

Bulk sequencing data Aligned reads Sample 2 Sample 1 (S2) (S1) S1 S2 S1 � � 0 . 5 0 . 17 0 . 33 0 . 17 0 S2 0 . 5 0 0 . 25 0 . 25 0 . 25 VAF matrix F (# variant reads / # total reads) Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 6 / 24

ISA ~3 billion base pairs Infinite Sites Assumption (Kimura 1969) No position in the genome mutates more than once. Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 7 / 24

AncesTree (El-Kebir et al. 2015) S1 � 0 . 5 � 0 . 17 0 . 33 0 . 17 0 S2 0 . 5 0 0 . 25 0 . 25 0 . 25 Ancestry graph (AG) Clonal trees VAF matrix F Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 8 / 24

AncesTree (El-Kebir et al. 2015) S1 � 0 . 5 � 0 . 17 0 . 33 0 . 17 0 S2 0 . 5 0 0 . 25 0 . 25 0 . 25 Ancestry graph (AG) Clonal trees VAF matrix F Observation Possible clonal trees ≡ AG spanning trees satisfying the sum condition : � F ij ≥ ∀ i ∈ { 1 , . . . , s } . F ik k child of j Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 8 / 24

AncesTree (El-Kebir et al. 2015) S1 � 0 . 5 � 0 . 17 0 . 33 0 . 17 0 S2 0 . 5 0 0 . 25 0 . 25 0 . 25 Ancestry graph (AG) Clonal trees VAF matrix F Variant Allele Frequency Factorization Problem (VAFFP) Given: VAF matrix F . Find: Usage matrix U and clonal matrix B such that F = 1 2 UB . Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 8 / 24

⇒ ⇒ Outline Goal: Find likely clonal tr Background 1 Appr 0.99 0.99 Methods 2 • Enumeration VAFFP • Noise in sequencing data 0.87 Handling noise 0.34 0.73 Shrinking the search space • 0.01 Results 3 � Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 9 / 24 Illumina sample .

E-VAFFP T G ( F ) F { … } � 0 . 5 � S1 0 . 17 0 . 33 0 . 17 0 , , S2 0 . 5 0 0 . 25 0 . 25 0 . 25 Enumeration VAFFP Given: VAF matrix F . Find: The set T ( G F ) of all ancestry graph spanning trees that satisfy the sum condition. How: Modified version of (Gabow and Myers 1978) Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 10 / 24 �

E-VAFFP T G ( F ) F { … } � 0 . 5 � S1 0 . 17 0 . 33 0 . 17 0 , , S2 0 . 5 0 0 . 25 0 . 25 0 . 25 Enumeration VAFFP (strict) Given: VAF matrix F . Find: The set T ( G F ) of all ancestry graph spanning trees that satisfy the sum condition. How: Modified version of (Gabow and Myers 1978) Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 10 / 24 �

Sources of noise DNA fragments Aligned reads Mutations Short reads …GATTACA… 0.75 0.25 Variant Sequencing Alignment 0.50 Calling 0.25 0.25 Output: VAFs Input: Mixed cell sample Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 11 / 24

Relaxed sum condition ✓ ! � F ij ≥ F ik ∀ i ∈ { 1 , . . . , s } k child of j Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 12 / 24

Relaxed sum condition ✓ ! ✓ � F ij + ε ≥ F ik ∀ i ∈ { 1 , . . . , s } k child of j Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 12 / 24

⇒ ⇒ Approximate ancestry graph Goal: Find likely clonal tr Appr 0.99 0.99 • • 0.87 0.34 0.73 • 0.01 1 Complete weighted digraph 2 Posterior robability of ancestry: beta-binomial model (El-Kebir et al. 2015) � 3 Enumerate spanning trees in weight order (Camerini et al. 1980) Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 13 / 24 Illumina sample .

Partial transitive reduction Goal: simplify ancestry graph Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 14 / 24

Partial transitive reduction 2-transitive 3-transitive Goal: simplify ancestry graph Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 14 / 24

Partial transitive reduction 2-transitive 3-transitive Goal: simplify ancestry graph k -PTR Subgraph resulting from removing all ≥ k -transitive edges. Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 14 / 24

Partial transitive reduction 3-PTR 2-transitive 3-transitive Goal: simplify ancestry graph k -PTR Subgraph resulting from removing all ≥ k -transitive edges. Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 14 / 24

Partial transitive reduction 2-transitive 3-transitive Goal: simplify ancestry graph k -PTR Subgraph resulting from removing all ≥ k -transitive edges. Tomlinson and Oesper (Carleton College) Tumor Phylogeny Inference Dec. 4, 2018 14 / 24