AimsMicrodosimetric indicators are commonly used to describe the energy deposition characteristics of ionizing radiation at the subcellular and DNA levels to predict radiation-induced biological effects. Monoenergetic electrons and beta-emitting radionuclides (β nuclides) have a certain range in low-density media, and their microdosimetric indicators in the cell nucleus are of great significance for understanding the characteristics of different types of β nuclides and the evaluation of relative biological effectiveness (RBE). MethodsMicrodosimetry in the nucleus is greatly affected by cell morphology, so in this paper, based on laser confocal tomography images of BEAS-2B and L-O2 cell lines, two types of mesh-type cell models were constructed to restore the morphological features of real cells. The Geant4-DNA toolkit was used to simulate the lineal energy and dose distribution of monoenergetic electrons (0.1 keV-1 MeV) and 5 types of beta nuclides (Cs-137, Cs-134, C-14, I-131, and H-3) in the cell nucleus with different initial energies. The effects of factors such as initial energy, sensitive site (radius = 2 nm, 10 nm, 30 nm, 300 nm), and cell type on microdosimetric quantities in the nucleus and their predictions of nuclide RBE were analysed. ResultsThe results show that for the same type of nuclide, there are different RBE values between BEAS-2B and L-O2 cells, with Cs-134 having a difference of 17.75% and H-3 having a difference of 10.58%. In addition, the mean lineal energy in the nucleus fluctuates with changes in the initial energy and sampling site radius of monoenergetic electrons, especially for low and medium energies where the sampling site radius has a significant effect. Moreover, the proximity functions of monoenergetic electrons with an initial energy below 1 keV are greatly affected by the cell shape, and those above 20 keV are greatly affected by the model volume. ConclusionsThe study confirms the necessity of using mesh cell models for microdosimetric analysis in the cell nucleus and predicting its biological effects by combining mesh models of real human cells to simulate microdosimetric indicators of monoenergetic electrons and beta-emitting nuclides. The estimation of the RBE of β nuclides achieved in this study has reference value for predicting radiobiological damage in different cell types.