Enrichment of nano delivery platforms for mRNA-based nanotherapeutics

Lipid-based nanoparticles(LNP)have shown significant progress in delivering mRNA for therapeutics,particularly with the success of coronavirus disease 2019(COVID-19)vaccines.However,there are still challenges,such as organ-specific targeting,sustained protein expression,immunogenicity,and storage that need to be addressed.Therefore,there is interest in developing additional nano drug delivery systems(DDS)to complement LNP technology.Some of these include polymer,lipid-polymer hybrid,organic/inorganic hybrid nanostructure,and inorganic nanoparticle.In our opinion,LNP technology may not be suitable for every disease scenario in categories such as infection disease,cancer,pulmonary disease,autoimmune disorders and genetic rare disease(among others).This is because different diseases may require distinct administration routes,doses,and treatment durations,as well as considerations for biological barriers that may lower the efficacy and/or exert safety concern.In this perspective,we will highlight the need and potential for enhancing the diversity of nano delivery platforms for mRNA-based nanotherapeutics.

LNP模型中的神经元滤波特征提取

目的 LNP(linear-nonlinear-Poisson)模型很好地解译了神经元的响应过程,其重要环节之一是线性滤波器的提取。针对传统i STAC(information-theoretic spike-triggered average and covariance)算法运用于LNP模型时的神经元特性表征不足、运动特征提取效果不佳等问题,特别是在处理低维度刺激问题时,提出了一种改进的i STAC神经元滤波特征提取算法。方法引入非触发刺激的统计量,从而更加准确地构建神经元滤波特征子空间的目标函数,同时增强系统的抗噪能力;采用变尺度法最大化目标函数,从而优化解空间,提升算法的收敛速率。结果不同非线性条件下对线性滤波器的恢复实验结果表明,新算法相较于传统i STAC算法在高维度刺激时保持较好的表征特性,在刺激维度小于6 500时有明显改善,且总体上优于STA(spike-triggered average)和STC(spike-triggered covariance)算法。结论提出的新算法适用范围更广,鲁棒性更强,能够运用于建立完整的基于视觉特性的视频运动特征提取模型。