Hepatocellular　carcinoma　(HCC)　is　a　heterogeneous　disease　and　the　second　most　common　cause　of　cancer-related　death　worldwide.　Marked　developments　in　genomic　technologies　helped　scientists　to　understand　the　heterogeneity　of　HCC　and　identified　multiple　HCC-related　molecular　subclasses.　An　integrative　analysis　of　genomic　datasets　including　196　patients　from　The　Cancer　Genome　Atlas　(TCGA)　group　has　recently　reported　a　new　HCC　subclass,　which　contains　three　subgroups　(iCluster1,　iCluster2,　and　iCluster3).　However,　the　transcriptional　molecular　characteristics　underlying　the　iClusters　have　not　been　thoroughly　investigated.　Herein,　we　identified　a　more　aggressive　subset　of　HCC　patients　in　the　iCluster1,　and　re-clustered　the　TCGA　samples　into　novel　HCC　subclasses　referred　to　as　aggressive　(Ag),　moderate-aggressive　(M-Ag),　and　less-aggressive　(L-Ag)　subclasses.　The　Ag　subclass　had　a　greater　predictive　power　than　the　TCGA　iCluster1,　and　a　higher　level　of　alpha　fetoprotein,　microscopic　vascular　invasion,　immune　infiltration,　isocitrate　dehydrogenase　1/2　mutation　status,　and　a　worse　survival　than　M-Ag　and　L-Ag　subclasses.　Global　transcriptomic　analysis　showed　that　activation　of　hedgehog　signaling　in　the　Ag　subclass　may　play　key　roles　in　tumor　development　of　aggressive　HCC.　GLI1,　a　key　transcriptional　regulator　of　hedgehog　signaling　upregulated　in　the　Ag　subclass,　was　correlated　with　poor　prognosis　of　HCC,　and　may　be　a　potential　prognostic　biomarker　and　therapeutic　target　for　Ag　subclass　HCC　patients.