
Our Perspective "Mapping the Transcriptome - realising the full potential of spatial data analysis" is out today in @CellCellPress cell.com/cell/fulltext/S0092…. A short thread on this work follows...

Senior Lecturer in Bioinformatics @UniofNewcastle, Director of Research @NclBNS. #rstats, Python, etc. Gateshead/Newcastle resident. Husband. Dad. Views my own.

Our Perspective "Mapping the Transcriptome - realising the full potential of spatial data analysis" is out today in @CellCellPress cell.com/cell/fulltext/S0092…. A short thread on this work follows...

Me too :) #UKGS24
This tweet is unavailable

So, Conservatives pledge to increase apprenticeships by 100k by cutting 1 in 8 degrees. I feel a little thread is in order.

How much more money do we need to dump into automating analysis for non experts until we realize that we need to better train scientists to interpret results!
Automating defaults rarely works. The HCA spend tens of millions on it and failed. Methods do not need to be fully automated and free of user intervention. Users need to be smart and know what they are doing


This is a massive problem @NCBI. As @TrostLab mentions, @NU_ITservice maintain this is a server-side problem, some clarity (and a fix!) would be very welcome:
Dear @NCBI @NLM_NIH in our institution @StudentsNCL in the UK we cannot access Pubmed or any other NIH/NCBI website anymore. this has been intermittently going on for months! Out IT says it comes from your side. can you please look into this! thanks a lot!

In the latest issue! Mapping the transcriptome: Realizing the full potential of spatial data analysis dlvr.it/T0VhqM

A PhD studentship available with me and collaborators @LabPeffers and @IGM_YoungLab from October 2024. findaphd.com/phds/project/bb… Combining omics data to identify regulators of chondrogenesis. Funded via @NLD_BBSRCDTP Deadline to apply 15th January #bioinformatics #PhD

Paper now officially published: cell.com/cell/fulltext/S0092…

Things you never thought you’d see in your career… featured content in Cell from “Cockell and colleagues”

Bringing 2023 to a close, the new issue features a 3D atlas of the developing human head, how the ineffective control of EBV-induced autoimmunity increases the risk for multiple sclerosis and a machine learning architecture that integrates cell types across single-cell datasets! To read more of our new issue click👉 cell.com/cell/current

Now online! Mapping the transcriptome: Realizing the full potential of spatial data analysis dlvr.it/SzrT2M

To quote ourselves, these features mean that the analysis of these types of data should be "spatial by default", and encourage inter-disciplinary collaboration with those in the geographical sciences to help transfer decades of knowledge from that domain

3) spatial heterogeneity - the relationships between variables is not static across space, and local relationships rarely look the same as the global picture

2) spatial autocorrelation - near values are more related to one another than distant ones. This seemingly obvious phenomena has a big effect on the assumptions of downstream statistics, particularly variable independence

We demonstrate that key features of spatially-resolved data can be shown in biological data (using Visium data by way of example). 1) The modifiable areal unit problem - that the scale you chose for your analysis can strongly influence the results of that analysis