How to count a place in a BP sequence as a "Epigenetic mark" or not
19 views
Skip to first unread message
Dan Kolis
Mar 11, 2024, 11:28:13 AMMar 11
to DIYbio
Ron B said:
> I was trying to picture some type of really unique cell, like a pancreas cell and wondering how many unique types of things it can do and produce. So I guess, on average, a cell would have about 500 unique functions. Have I missed the point completely, again? "
Dan K says:
Your open question is incredibly important. I would say there is no number answer until a heap of situational issues are defined...
Proposed "it counts" status:
*) Place is in a gene which can be expressed > 1 for 50% probability within any 500 hours of linear life. If Mitosis is slower then this then the interval is 5% of species lifespan. Baseline Num is queisent life, not "Birth phases", but to clarify, many species require 2 numbers per species.... per cell type.
*) Counts either "On" or "Off" with any near covalent mod, like CH2 or -OH ( There are others ). Usually.
*) Counts even if always preloaded from ancestor cell(s).
*) Any expression 'counts' if it either leaves a nucleus or creates a change in another gene expression.
*) Near precise alleles, do not count. Identical Mat or Pat of course, counts as one.
*) Does not count if it never, ever seems to 'do anything' to any transcription at all.
*) Counts even if its an Intron; "probably".
*) Does Not count within 2K BP of Telomere or Centromere.
Two numbers are reportable for any cell type for species like Mammals with birthing, moving to quiescent life. -EpMarkN is pre partum, +EpMarkN is post partum count.
So sure 500 is as good as any other I assume you mean +EpMarkN for Homo S. for lets say: Abs average for all cell types. Weighting is per "1 per issue type", not per cell count... Or Pancreas, sure..
Re-Published 'here and there", on WWW Google share spot(s) !
Well, uhh nobody knows enough to state numbers I suspect. if they do they should tolerance the estimate and hoping for 25 % would probably be a as good as it gets.
If anyone is interested in tightening up the specification and/or some programs to try it one real stuff post it here, please.
from Daniel Kolis and Ronald Brooks
11 Mar 2024
> I was trying to picture some type of really unique cell, like a pancreas cell and wondering how many unique types of things it can do and produce. So I guess, on average, a cell would have about 500 unique functions. Have I missed the point completely, again? "
Dan K says:
Your open question is incredibly important. I would say there is no number answer until a heap of situational issues are defined...
Proposed "it counts" status:
*) Place is in a gene which can be expressed > 1 for 50% probability within any 500 hours of linear life. If Mitosis is slower then this then the interval is 5% of species lifespan. Baseline Num is queisent life, not "Birth phases", but to clarify, many species require 2 numbers per species.... per cell type.
*) Counts either "On" or "Off" with any near covalent mod, like CH2 or -OH ( There are others ). Usually.
*) Counts even if always preloaded from ancestor cell(s).
*) Any expression 'counts' if it either leaves a nucleus or creates a change in another gene expression.
*) Near precise alleles, do not count. Identical Mat or Pat of course, counts as one.
*) Does not count if it never, ever seems to 'do anything' to any transcription at all.
*) Counts even if its an Intron; "probably".
*) Does Not count within 2K BP of Telomere or Centromere.
Two numbers are reportable for any cell type for species like Mammals with birthing, moving to quiescent life. -EpMarkN is pre partum, +EpMarkN is post partum count.
So sure 500 is as good as any other I assume you mean +EpMarkN for Homo S. for lets say: Abs average for all cell types. Weighting is per "1 per issue type", not per cell count... Or Pancreas, sure..
Re-Published 'here and there", on WWW Google share spot(s) !
Well, uhh nobody knows enough to state numbers I suspect. if they do they should tolerance the estimate and hoping for 25 % would probably be a as good as it gets.
If anyone is interested in tightening up the specification and/or some programs to try it one real stuff post it here, please.
from Daniel Kolis and Ronald Brooks
11 Mar 2024
Reply all
Reply to author
Forward
0 new messages