Sarcomas represent a heterogeneous group of mesenchymal malignancies arising at various locations in the soft tissue and bone. Though a rare disease, sarcoma affects ~200,000 patients worldwide every year. The prognosis of patients with sarcoma is poor, and targeted therapy options are limited; therefore, accurate diagnosis and classification are essential for effective treatment. Sarcoma samples were acquired from 199 patients, in which
TP53
(39.70%, 79/199),
CDKN2A
(19.10%, 38/199),
CDKN2B
(15.08%, 30/199),
KIT
(14.07%, 28/199),
ATRX
(10.05%, 20/199) and
RB1
(10.05%, 20/199) were identified as the most commonly mutated genes (>10% incidence). Among 64 soft-tissue sarcomas that were unclassified by immunohistochemistry, 15 (23.44%, 15/64) were subsequently classified using next-generation sequencing (NGS). For the most part, the sarcoma subtypes were evenly distributed between male and female patients, while a significant association with sex was detected in leiomyosarcomas. Statistical analysis showed that osteosarcoma, Ewing's sarcoma, gastrointestinal stromal tumors and liposarcoma were all significantly associated with the patient age, and that angiosarcoma was significantly associated with high tumor mutational burden. Furthermore, serially mutated genes associated with myxofibrosarcoma, gastrointestinal stromal tumor, osteosarcoma, liposarcoma, leiomyosarcoma, synovial sarcoma and Ewing's sarcoma were identified, as well as neurotrophic tropomyosin-related kinase (
NTRK)
fusions of
IRF2BP2-NTRK1, MEF2A-NTRK3
and
ITFG1-NTRK3
. Collectively, the results of the present study suggest that NGS-targeting provides potential new biomarkers for sarcoma diagnosis, and may guide more precise therapeutic strategies for patients with bone and soft-tissue sarcomas.