To the Editor: Assignment of structure and function to all genes and gene products (such as proteins) of all organisms represents a major challenge in this postgenomic era. Here we present the I-TASSER Suite (http://zhanglab. ccmb. med. umich. edu/I-TASSER/download/), a stand-alone software package for protein structure and function modeling.

The gap between the number of proteins with known sequences and the number of proteins with experimentally characterized structure and function keeps increasing. One way to narrow this gap is by developing advanced computational approaches for modeling structure and function from sequences, where progress has been recently witnessed in community-wide blind experiments1, 2. I-TASSER3 was originally designed for protein structure modeling by iterative threading assembly simulations. It was recently extended for structure-based function annotation by matching structure predictions with known functional templates4, 5. Here we introduce the I-TASSER Suite, a stand-alone package implementing the I-TASSER–based protein structure and function modeling pipelines. Although the on-line I-TASSER server is established and widely used in the community, limited computing resources from a single laboratory have prevented large-scale applications of these algorithms. We expect that the development of the stand-alone package will remove the computing resource barriers and therefore enable benchmarking of new structure and function modeling methods. The I-TASSER Suite pipeline consists of four general steps: threading template identification, iterative structure assembly simulation, model selection and refinement, and structure-based function annotation (Fig. 1a). In the first step, the query is threaded by LOMETS through a nonredundant structure library to identify structural templates. LOMETS is a meta-threading method containing eight fold-recognition programs (PPAS, Env-PPAS, wPPAS, dPPAS, dPPAS2, wdPPAS, MUSTER and wMUSTER). These programs are generally based on sequence profile-toprofile alignments, but with various structural features combined (Supplementary Methods). Such variation is important for generating complementary alignments, which increase the coverage of template detections.