josh-atkins.bsky.social
@josh-atkins.bsky.social
A shout out to @CRUKresearch and everyone that donates/volunteers to them. Without funding this kind of work isn't possible. Also to everyone that donated samples in the @uk_biobank. The research community cannot thank you enough for bulding this wonderful resource
January 21, 2025 at 2:09 PM
In a non-scientific summary -> we have shown proteins in the blood associated with changes before cancer diagnosis and in some cases greater than a decade out. Many of these are also driven by genetics that are associated with cancer risk. Paving the way for precision prevention
January 21, 2025 at 2:04 PM
We found 4 proteins-cancer associations that had evidence from all analysis performed
January 21, 2025 at 1:59 PM
Out of the 618 - we wanted to identify if there was any genetic evidence to support these associations -> We used cis-pQTL but also created exome scores (projected into the full ukbb exome data). We found strong evidence for one cis hit and 28 proteins using the exome-scores
January 21, 2025 at 1:53 PM
Out of the 618, we wanted to see if any were targets of drugs currently available - we found 38 drugs including Pembrolizumab for PDCD1. This work was done by the very talented @uzor_md
January 21, 2025 at 1:48 PM
We identified 182 proteins that differed <3 years before diagnosis --> providing a potential marker for immediate cancer
January 21, 2025 at 1:43 PM
when we stratified out into < 3 yrs to diagnosis, 3-7 yrs and > 7 yrs (median time do diagnosis in this group was 12 yrs). In the >7 yrs group we identified 107 protein-cancer associations.....
January 21, 2025 at 1:37 PM
Using the invaluable @uk_biobank and the @OlinkProteomics data, we investigated 1,463 proteins across 19 common cancers, finding 618 protein-cancer associations.... (Thanks Trishna Desai for this great figure 1)
January 21, 2025 at 1:32 PM