Organ representation of Cancer Genecorpus 14M

#430
by lurcelay - opened

Hi! Would it be possible to share the organ distribution of the 14M cancer samples as it has been done with Genecorpus 103M in the paper? It would be very valuable in order to asses the performance of the cancer-tuned Geneformer in different cancer types.

image.png

Thank you for your question!:
brain 21.84%
immune 20.72%
breast 11.75%
large intestine 6.51%
lung 5.18%
liver 4.66%
kidney 4.27%
bone marrow 4.11%
pancreas 3.53%
skin 2.24%
stomach 1.92%
adipose 1.84%
eye 1.82%
muscle 1.79%
lymph node 1.74%
reproductive system 1.16%
abdomen 1.10%
peripheral nervous system 1.02%
nasal 0.92%
spinal cord 0.90%
nervous system unspecified 0.34%
unlabeled 0.29%
prostate 0.19%
small intestine 0.07%
adrenal 0.03%
mouth 0.03%
bone 0.01%

ctheodoris changed discussion status to closed

Sign up or log in to comment