CIBERSORT is a computational method for characterizing cell composition in complex tissues from gene expression profiles. It outperforms other methods in handling noisy data, unknown mixture content, and closely related cell types. The method uses linear support vector regression (SVR) to estimate the relative fractions of different cell types, with an additional feature selection step to adaptively select genes from a reference signature matrix. CIBERSORT was validated using leukocyte gene signatures and showed high accuracy in deconvolving leukocyte fractions in bulk tumor samples, peripheral blood, and formalin-fixed, paraffin-embedded (FFPE) specimens. It also demonstrated superior performance in detecting rare cell types and resolving closely related cell types. CIBERSORT is available at <http://cibersort.stanford.edu>.CIBERSORT is a computational method for characterizing cell composition in complex tissues from gene expression profiles. It outperforms other methods in handling noisy data, unknown mixture content, and closely related cell types. The method uses linear support vector regression (SVR) to estimate the relative fractions of different cell types, with an additional feature selection step to adaptively select genes from a reference signature matrix. CIBERSORT was validated using leukocyte gene signatures and showed high accuracy in deconvolving leukocyte fractions in bulk tumor samples, peripheral blood, and formalin-fixed, paraffin-embedded (FFPE) specimens. It also demonstrated superior performance in detecting rare cell types and resolving closely related cell types. CIBERSORT is available at <http://cibersort.stanford.edu>.