| | |

Opencell

Discussion by Manu Leonetti at Chan Zuckerberg hub.

Self driving cars are taught in simulation. Similar biology should be also done as simulations. The cell needs to be thought as devices. How does information flow within a cell and do that in cells. Use GFP to tag the cells but also later use it for finding protein – protein interaction.

Opencell.czbiohub.org

organelle.czbiohub.org

This is a wiring diagram of the proteome. These experiments take a long time to do…since they are done 1 t a time. Coarse graining cellular representations with organelle profiling: mesoclale polifilng ath the organelle level.

The organelle becomes biomarkers of the cellular state.

Later do morphological analysis of whole organelle by Mass Spectrometry…

Pan-cellular organelle collection. This generalizes 30 different protein markers for different organelles.

Human stem cell generalization is in progress.

Analysis by Graph based analytical strategy (cf scRNASeq). Clusters correspond to organelles. This is UMAP version that can be browsed on the web and can show a graph with relationships and complexes and proteins that are next to each other are close together on the graph.

This can be seen remodeled in a OC43B-coronavirus infection. (infection vsn noon-infected).

Cells either move things around or by abundance changes. Subcellular remodeling reveals cellular responses not captured by abundance changes. They found link between ferroptosis and viral infection.

This data is AI ready.

czbiohub-sf/grassp. MAX FRANK. AI-ready subcellular proteomics. All the code is on github.

How is cell a basic unit of biological computation – They are non-linear dynamic system. Need to know the starting state – to use dynamic data.

In imaging – interpretability is hard (Imagine if we could watch cells think).

Multi-modality – imaging as a foundation for all the omics. Project all the molecular information to live cells.

AI interpreter for images that is learnt from omics embeddings.

High throughput optical screening platform…

How do you build generalizability? How do we go from cells → tissues ?

Imaging on the RNA molecules may also be possible.

Similar Posts

  • |

    Reinforcement learning

    Reinforcement learning is a method that drives learning and memory in primitive species such as birds, humans and other living species to its use in machine learning. It is used to influence the behavior of us humans on social media to its use to train machines. The essential components were initiated by BF Skinner 20th…

  • Detection of AI created content

    As AI has progressed generating code, writing, music or other intelligent language based skills through an LLM there is a parallel growth in detection of AI generated content.  Like everything AI, it is a probability game. Trying to estimate the combination of words/tokens and comparing with what is in the model. Using a standard model…

  • Bug eyes with camera

    It is known that eyes of the common fly is a compound eye with multiple detectors and multiple lenses. It has always been thought that the 180 degree vision would be very useful and could be very high resolution. Though interesting to think about in terms of camera’s, it has been difficult to create multiple…

  • |

    ALS and TDP43

    Almost 90% of ALS is thought to be sporadic. The reasons are not well understood and thought to be oxidative stress, damaged endoplasmic reticulum, mitochondria, cytoskeleton or misregulation of RNA pathways. One protein implicated is TDP43 which sits as a junction of ALS pathogenic path by binding to RNA modulation synthesis, splicing, stability and transport….