Machine learning-assisted high-content analysis of pluripotent stem cell-derived embryos in vitro

Area of Science:

• Developmental Biology
• Stem Cell Biology
• Bioinformatics

Background:

• Stem cell-based embryo models offer novel platforms for studying early postimplantation development.
• Induced pluripotent stem cells (iPSCs) can form embryo models resembling in vivo gastrula stages.

Purpose of the Study:

• To establish an efficient and unbiased machine learning workflow for analyzing stem cell-based embryo models.
• To identify factors that promote the morphogenesis of these embryo models.

Main Methods:

• Utilized a high-content screening system to collect 3D images of iPSC-derived embryo models (ITS embryos).
• Developed a machine learning workflow for multi-dimensional feature extraction and quantification.
• Performed high-content screening of small molecules and cytokines.

Main Results:

• Demonstrated variability in embryo-like structure formation among different pluripotent stem cell (PSC) lines.
• Identified BMP4 as a key factor that significantly promotes ITS embryo morphogenesis.
• Established an innovative analytical strategy for stem cell-based embryo models.

Conclusions:

• The study presents a novel machine learning approach for analyzing stem cell-based embryo models.
• BMP4 plays a crucial role in promoting the development of these in vitro embryo models.
• This work advances the understanding of early embryonic development and stem cell applications.