Present day biology generates a wealth of data from which tumor biology and tumor evolution can be inferred, and therapeutic strategies can be developed. Most clinical (and research) samples are comprised of a mixture of different cells, including multiple different populations of cancer cells (‘subclones’) and a variety of normal cell types. Effective use of these data thus requires data from individual cell types to be deconvolved. A number of subfields are independently developing mathematical and statistical deconvolution techniques. While at first glance these subfields are largely disparate, there are clear opportunities for synergies to be developed. This workshop will fill this gap by bringing together researchers from several of these fragmented subfields across a wide range of backgrounds (computational biology, mathematics, statistics, computer science, biomedicine, etc.) to provide a synergistic forum for cross-disciplinary learning, discussion and collaboration.