Myelodysplastic syndrome (MDS) is a group of myeloid neoplasms with hematopoietic stem cell origin. Some patients with MDS have a natural course of “refractory anemia (RA)/refractory thrombocytopenia with multilineage developmental abnormalities (RCMD) → refractory anemia with primitive cellular excess-1 (RAEB-I) → refractory anemia with primitive cellular excess (RAEB-H) → secondary acute myeloid leukemia (sAML)”. → secondary acute myeloid leukemia (sAML)” is a natural disease progression, which is an excellent disease model to study the evolution of malignant clones of myeloid neoplasms. In recent years, several large series of whole genome or targeted gene sequencing studies in MDS patients have been completed, which initially revealed the molecular basis of MDS pathogenesis. 1. Genes involved in MDS patients There are about 60 genes involved in MDS, which are divided into the following major categories: (1) RNA shearing, such as SF3B1, SRSF2, U2AF1, ZRSR2, etc.; (2) DNA methylation, such as TET2, DNMT3A, IDH1/IDH2, etc.; (3) chromatin remodeling, such as ASXL1, EZH2, etc.; (4) transcription factors, such as (4) transcription factors, such as RUNX1, BCOR, etc.; (5) DNA repair, such as p53, etc.; (6) adhesion factors, such as STAG2, etc.; (7) RAS signaling pathways, such as CBL, NRAS, KRAS, NF1, etc. The most frequently affected genes are SF3B1, TET2, SRSF2, ASXL1, DNMT3A and RUNX1, all of which have a mutation frequency of 10% or more. (1) As long as the diagnostic criteria for MDS are met, mutations are present even if the bone marrow primitive cells are zero; (2) Most patients have two or more mutations, and the number of mutations increases with RA→RCMD→RAEB-I→RAEB-II; (3) The mutation spectrum of MDS is different from that of primary AML. (3) The mutation spectrum of MDS is different from that of primary AML, further confirming at the molecular level that MDS and AML are two different disease entities, although MDS was previously referred to as “pre-leukemia”; (4) The genes encoding RNA shear subunits and DNA methylation regulatory genes may be the starting mutations for malignant clones of MDS, while other mutations are mainly involved in subclonal evolution. 2, mutations and MDS diagnostic typing The relationship between mutations and some clinical parameters has been initially discussed, such as TET2, RUNX1, CBL and NRAS mutations are associated with increased bone marrow primitive cells, TET2, CBL and NRAS mutations are associated with increased peripheral blood mononuclear cells, NRAS, p53, RUNH mutations are associated with decreased peripheral blood platelet count, p53 mutations are associated with complex karyotypes, and the relationship between gene mutations and morphological alterations in abnormal development of granulocyte, red and megakaryocyte lineages will be the focus of the next studies. Although cytogenetic and/or genetic abnormalities are present in about 90% of MDS patients, meaning that clonal evidence can be found, the only existing genetic abnormality in MDS that has been found to be more specifically correlated with myeloid ringed iron granulocytosis is the SF3B1 gene mutation, which can be used for staging diagnosis, and about 70% of RAS and RCMD-RS have SF3B1 gene mutations. 3. Mutations and prognosis The available results confirm that, similar to chromosomal abnormalities, the overall survival of patients becomes worse as the number of mutations increases. Prognostic analysis of individual mutations showed that SF3B1 mutations had a good prognosis, while mutations in SRSF2, U2AF1, DNMT3A, ASXL1, EZH2, RUNX1, CBL, NRAS, and KRAS suggested a poor prognosis. New prognostic score systems have been proposed by combining gene mutations with other prognostic parameters of MDS, but these systems need further validation. The development of treatment strategies for MDS is mainly based on the prognostic grouping of the international prognostic score system for the disease itself, as well as on some factors of the patients themselves, such as age, general condition score and risk grouping for comorbid diseases, combined with the subjective wishes of the patients. There are generally accepted models for predicting the efficacy of erythropoietin and immunosuppression, but there are no reliable biological markers for predicting the efficacy of the epigenetic drugs azacitidine and decitabine, which have been used as first-line treatment options. However, because of the small number of cases, large series of rigorously designed clinical trials are needed to confirm this. Although the research on the molecular basis of MDS pathogenesis has made great progress in recent years, the following aspects need to be addressed and strengthened: First, there are many genes involved in MDS, and there is still a bottleneck in health economics as a routine clinical test, in addition, the incidence of some mutations is extremely low and their clinical significance is not clear, therefore, at this stage, at least SF3B1, TET2, SRSF2 Second, little is known about the specific mechanisms of the identified genetic abnormalities in the occurrence, development and evolution of MDS, especially how the genes act in concert with each other. The next step of research will be to explore the clonal evolution of different subtypes and subgroups of patients with different cytogenetic abnormalities, and through the establishment of in vitro transgenic mouse models, it is expected that the molecular mechanism of MDS pathogenesis will be truly analyzed at the molecular level; thirdly, it has been confirmed that there are multiple subclones with different combinations of mutations in MDS patients. Is the relapse after complete remission a recurrence of the original clone? Furthermore, are the subclones with different malignancies differently sensitive to different drugs? Only the answers to questions such as these will make it possible to propose an individualized treatment plan based on molecular abnormalities.