RNA-Seq studies indicate that the cancer transcriptome are shaped by genetic changes, variation in gene transcription, mRNA processing, editing and stability, and the cancer microbiome. Deciphering this variation and understanding its implications on tumorigenesis requires sophisticated computational analyses, and being able to tackle analyses of bulk RNA-Seq as well as transcriptomes of individual tumor cells. Most RNA-Seq analyses rely on methods that first map short reads to a reference genome, and then compare them to annotated transcripts or assemble them. However, this strategy can be limited when the cancer genome is substantially different than the reference or for detecting sequences from the cancer microbiome. `Assembly first' (de novo) methods that combine reads into transcripts without any mapping are a compelling alternative. The assembled transcriptome can then be used to identify mutations, fusion transcripts, splicing patterns, expression levels, tumor-associated microbes, and – if collected from single cells – characterize tumor heterogeneity. There is thus an enormous need for computationally efficient, accurate and user friendly tools for transcriptome reconstruction and analysis in cancer. Trinity, first released in mid-2011 and freely available as Open Source, is the leading software for de novo RNA-Seq assembly, executed millions of times by thousands of rsearchers, over 4k literature citations, and now includes a host of modules for downstream analyses, contributed by the Trinity development team or contributed by 3rd party developers. Here, we will continue to enhance and maintain Trinity and further develop our Trinity Cancer Transcriptome Analysis Tookit (CTAT) as leading tool suite for bulk and single-cell cancer transcriptomics. We will tailor analytic modules for critical tasks in cancer biology, working with a network of cancer researchers on Driving Cancer Projects (Aim 1). We will continue to update the Trinity software to enhance the core algorithm, leveraging new sequencing technologies and integrating genome data with genome-free assembly (Aim 2). We will integrate Trinity CTAT into the NCI cloud computing platform via FireCloud for scalable cancer transcriptome data processing and analyses (Aim 3). We will grow the Trinity cancer user community, using online and in person training and support (Aim 4), to allow any cancer researcher to leverage it in diverse modalities.

The RNA in cancer cells can be altered in its sequence, level and structure, and understanding these alterations can help us diagnose disease and find new drug targets. Tools enabling comprehensive analyses for bulk and single-cell transcriptomes are essential for this. Here, we enhance and maintain the Trinity software and our Cancer Transcriptome Analysis Toolkit for this purpose, and build computational resources and training materials to help cancer researchers use it.