Major depressive disorder (MDD) is a complex psychiatric illness, with synaptic plasticity playing a key role in its pathology. Our study aims to investigate the molecular basis of MDD by analyzing synaptic plasticity-related gene expression at the single-cell level. Utilizing a published snRNA-seq dataset (GSE144136), we identified Excitatory.neurons_1 as the cell cluster most associated with MDD and synaptic plasticity through cell clustering, gene set enrichment analysis (GSEA), and pseudotime analysis. Integrating the bulk RNA-seq data (GSE38206), we identified CASKIN1 and CSTB as hub genes via differential expression analysis and machine learning methods. Further exploration of the relevant mechanisms was performed via cell–cell communication and ligand-receptor interaction analysis, functional enrichment analysis, and the construction of molecular regulatory networks, highlighting miR-21-5p as a key biomarker. We propose that elevated miR-21-5p in MDD downregulates CASKIN1 in Excitatory.neurons_1 cells, resulting in decreased neural connectivity and altered synaptic plasticity. As our analyzed snRNA-seq dataset consists solely of male samples, these findings may be male-specific. Our findings shed light on potential mechanisms underlying synaptic plasticity in MDD, offering novel insights into the disorder’s cellular and molecular dynamics.

Microbiota have emerged as a promising tool for estimating the post-mortem interval (PMI) in forensic investigations. The role of oral and nasal microbiota in cadaver decomposition is crucial; however, their distribution across human cadavers at different PMIs remains underexplored. In this study, we collected 88 swab samples from the oral and nasal cavities of 10 healthy volunteers and 34 human cadavers. Using 16S rRNA gene sequencing, we conducted comprehensive analyses of the alpha diversity, beta diversity, and relative abundance distribution to characterize the microbial communities in both healthy individuals and cadavers at varying PMIs and under different freezing conditions. Random forest models identified Firmicutes, Proteobacteria, Bacteroidota, Actinobacteriota, and Fusobacteriota as potential PMI-associated biomarkers at the phylum level for both the oral and nasal groups, along with genus-level biomarkers specific to each group. These biomarkers exhibited nonlinear changes over increasing PMI, with turning points observed on days 5, 12, and 22. The random forest inference models demonstrated that oral biomarkers at both the genus and phylum levels achieved the lowest mean absolute error (MAE) values in the training dataset (MAE = 2.16 days) and the testing dataset (MAE = 5.14 days). Additionally, freezing had minimal impact on the overall phylum-level microbial composition, although it did affect the relative abundance of certain phyla. At the genus level, significant differences in microbial biomarkers were observed between frozen and unfrozen cadavers, with the oral group showing greater stability compared to the nasal group. These findings suggest that the influence of freezing should be considered when using genus-level microbial data to estimate PMIs. Overall, our results highlight the potential of oral and nasal microbiota as robust tools for PMI estimation and emphasize the need for further research to refine predictive models and explore the environmental factors shaping microbial dynamics.

Although progress has been seen for low template DNA (LT-DNA) analysis in the last few years, the obtainment of full short tandem repeat (STR) profiles remains challenging in forensic genetics. Preamplification treatment could help increase the recovery of genetic information in STR typing. In this study, an efficient amplification method was investigated, which shows that the application of the abasic site in a single primer for targets could efficiently delete the primer-binding site of PCR products to achieve semi-linear amplification (abSLA PCR). Further, the position of abasic sites located at 8th to 10th from the 3’ end of the primers could facilitate PCR amplification. A significant increase in the recovery of STR loci could be obtained using a 4-plex STR pre-amplification (D8S1179, D21S11, D7S820, and CSF1PO) coupled to Identifiler Plus kit in low template of genomic DNA or single cells. Altogether, the application of abSLA amplification method can facilitate conventional STR typing in LT-DNA analysis.