Keynote speakers and Lectures

Charles Cantor (Agena Biosciences, Sequenom, Retrotope and Boston University, USA)


Tutorial: Is precision medicine possible? 

Research: Sensitive detection of low levels of cancer-specific DNA sequence difference

Philippe Gorry (GREThA UMR CNRS 5113, Universitè de Bordeaux, France)


Research Topic: “Economic Dimensions of Precision Medicine in Oncology”

Precision medicine is a new paradigm for disease treatment and prevention, taking in account genes, environment, and lifestyle for each patient. In oncology, it is the use of genomic information about person’s tumor in order to individualize treatments. It is rapidly gaining prominence in OECD countries, and it is supported by the pharma industry, moving from blockbuster to niche buster drug model, with the support of the Orphan Drug legislation. Today, the increased availability of targeted therapies with their high cost, and their use for chronic disease, raises a debate about accessibility, cost-effectiveness, and reimbursement. So, if precision medicine approaches become part of routine healthcare, paying for precision drugs would be a future challenge by checking traditional models of health protection systems. We will take advantage of on-going research on the Orphan Drug market to explore the economics dimensions of niche buster drug model on innovation dynamics, intellectual property protections, R&D return on investment and welfare impact.

Marcin Imielinsky (Weill Cornell Medical College, Cornell University, New York, USA)


Tutorial: Deciphering junctions, copy number, and walks in cancer whole genomes with linked-reads

 The altered structure of cancer DNA can be represented as a genome graph of nodes representing reference genomic intervals (DNA alleles) and edges representing junctions (3-5 phosphodiester bonds).    A subset of edges in the genome graph are variant– representing signed pairs of genomic locations that were previously distant (e.g. in the reference genome) and are made adjacent (e.g. in a cancer) through DNA breakage and fusion.  The copy number and rearrangement states of cancer genomes are constrained by junction balance, a constraint that arises from from the physical intuition that every copy of every (non-telomeric) interval in the genome must have a left and a right neighbor.   In practical terms, junction balance requires the copy number of nodes in the genome graph to be consistent with the copy number of incoming and outgoing edges. I demonstrate how this inference can be performed on standard cancer whole genome sequencing (WGS) data through the analysis of read depth subject to junction balance constraints.   We implement this robustly as a mixed-integer quadratic program optimization (JaBbA, which also allows for loose ends that represent missing junctions (e.g. rearrangements occurring in unamappable genomic regions) and result in copy number changes unaccounted for by graph topology.   I also present an R toolkit for analyzing, exploring, and visualizing annotated cancer genome graphs (gGnome,  One application of gGnome is decomposition of genome graphs into walks representing possible phases of rearranged alleles.   Using standard WGS, the phase of these alleles cannot be often unambiguously reconstructed at long distances (e.g. 100Kbp-1Mbp).   I demonstrate how the analysis of walks in the context of long-range sequencing data, such as linked reads arising from the barcoding and sequencing of short reads arising from high molecular weight (10-500 Kbp) DNA fragments via the 10X Chromium platform, can be leveraged to resolve long distance phases and loose ends.  These phased reconstructions of can be used to infer the consequences of complex structural variations and understand the evolution of complex rearranged loci across multiple tumor samples from individual patients.

Research: Highly prevalent noncoding somatic indel hotspots target lineage-defining genes in human cancer

Whole genome sequencing analysis of lung adenocarcinomas revealed indel hotspots in surfactant protein genes (SFTPA1, SFTPB, and SFTPC). Extrapolation to other solid cancers demonstrated highly recurrent and tumor-type-specific indel hotspots targeting the noncoding regions of highly expressed genes defining certain secretory cellular lineages: albumin (ALB) in liver carcinoma, gastric lipase (LIPF) in stomach carcinoma, and thyroglobulin (TG) in thyroid carcinoma. The sequence contexts of indels targeting lineage-defining genes were significantly enriched in the AATAATD DNA motif and specific chromatin contexts, including H3K27ac and H3K36me3. Our findings illuminate a prevalent and hitherto unrecognized mutational process linking cellular lineage and cancer. The findings may also represent a novel noncoding driver phenomenon that is under selection in multiple cancer types.

Bud Mishra (Courant Institute of Mathematical Sciences and Tandon School of Engineering, NYU, New York, USA)

Bud Mishra

Tutorial: “Learning Cancer Clocks

Research: “Genomics Technologies: Try Again! Fail Again!! Fail Better!!!”

As the genomic analysis of tumors and their usage in causality/progression analysis have gained momentum, we are frustrated by the problem of not being able to create data that reflect tumor’s heterogeneity (single cell/single molecule), genomic instability (resulting in aneuploidy, structural changes such as fusion), mobility (circulating tumor cells and cell free DNA), epigenetic, etc. We will describe a solution based on a new approach: SubOptical Mapping, which is ultra-cheap, fast (low latency/high throughput), accurate and potentially “highly disruptive.” The talk will focus on various questions related to data analysis and their applications: Image Analysis, Algorithms for Variant Detection, Haplotype Phasing, Whole Genome Off-target Activities (e.g., with CRISPR assays), Transcriptomics (isoforms, small copy number), Sequence Assembly and Clinical Genomics. Additional topics may be discussed upon request.

Eugenio Montini (San Raffaele Telethon Institute for Gene Therapy, HSR-TIGET, Milan, Italy)


Tutorial: “Vector marking as tool to study cancer clonality and anticancer drug resistance”

Topic: “Lentiviral vector mediated insertional mutagenesis for the discovery of novel culprits of anticancer drug resistance”

Quaid Morris (Banting and Best Department of Medical Research Terrence Donnelly Centre for Cellular and Biomolecular Research, University of Toronto, Canada)


Tutorial: “A user’s guide to subclonal reconstruction”

Research topic: “Reconstructing the evolutionary history of cancers using DNA-seq data”

Ben Raphael (Princeton University, Princeton, USA)


Tutorial: Identification of Cancer Genes, Pathways, and Networks

A key challenge in interpretation of cancer genome sequencing data is to distinguish driver mutations responsible for cancer development from random passenger mutations.  We will survey a number of computational approaches to predict driver mutations.  These include: (1) methods to identify recurrently mutated genes across cohorts of cancer genomes/exomes; (2) approaches that examine clustering of mutations in pathways and interaction networks; (3) algorithms that consider correlations between mutations including mutual exclusivity and co-occurrence.  We will demonstrate how these approaches have been used in a number of large-scale cancer sequencing projects. 

Research: Algorithms for Inferring Evolution and Migration of Tumors

Cancer is an evolutionary process driven by somatic mutations that accumulate in a population of cells that form a primary tumor.  In later stages of cancer progression, cells migrate from a primary tumor and seed metastases at distant anatomical sites.  I will describe algorithms to reconstruct this evolutionary process from DNA sequencing data of tumors.  These algorithms address challenges that distinguish the tumor phylogeny problem from classical phylogenetic tree reconstruction, including challenges due to mixed samples and complex migration patterns.

Andrea Sottoriva (The Institute of Cancer Research, London, UK)


Research: “Quantifying clonal selection in human cancer using next-generation sequencing data”

Recent studies have identified prevalent subclonal architectures within many cancer types. However, the temporal evolutionary dynamics that produce these subclonal architectures remain largely unknown. Although we can detect subclones in cancer, measuring what these subclones are doing remains challenging. We developed a computational model of tumour evolution based on branching processes that allows measuring the dynamics of clonal selection using high throughput sequencing data from bulk cancer cell populations. Application of our method to high-depth sequencing data from gastric and lung cancers revealed that detectable subclones consistently emerged early during tumour growth and had considerably large fitness advantages (>20% growth advantage). Our quantitative framework provides new insight into the evolutionary history of cancers by facilitating the measurement of fundamental evolutionary parameters in individual patients.

Tutorial: “Evolutionary modelling and Bayesian inference on cancer genomic data”

Genomics has revolutionized cancer research and has led to the generation of astonishing amounts of data. Such large-scale multidimensional datasets contain an overwhelming amount of information that is often hard to make sense of. In particular, there is the need to integrate genomic information into a solid mechanistic framework based on cancer evolution that allows understanding what has happened in individual patients over time (the big hidden variable in cancer). In evolutionary biology, mathematical modeling and in particular population genetics have helped making sense of genetic data for decades, however such wealth of literature is often underexploited in cancer. In this tutorial we present the use of established statistical inference frameworks, in particular Approximate Bayesian Computation, that allow integrating complex mathematical and computational models of tumour evolution with commonly available genomic data, offering unprecedented insight into the dynamics of cancer evolution.

Simon Tavarè (Cancer Research UK, Cambridge, UK)


Tutorial: “An introduction to Approximate Bayesian Computation”

Research: “How to study tumours in 3.5D”