Ewan Birney
@ewanbirney.bsky.social
Executive Director EMBL. I have an insatiable love of biology. Consultant to ONT and Cantata (Dovetail)
That's definitely a bad rendering (i don't know where it is from). It looks like a pretty sane DNA binding protein from the (quite a few) protein structures out there. I wouldn't take Wikipedia's choice of protein structure as the best choice (UniProt a far better aggregation point)
November 11, 2025 at 5:11 PM
That's definitely a bad rendering (i don't know where it is from). It looks like a pretty sane DNA binding protein from the (quite a few) protein structures out there. I wouldn't take Wikipedia's choice of protein structure as the best choice (UniProt a far better aggregation point)
Good morning Heidelberg!
November 5, 2025 at 6:13 AM
Good morning Heidelberg!
British electricity grid is on a real renewable (windy) roll this week and at the moment, during the day, we're exporting more electricity (to Norway, Ireland, Belgium and Denmark) than we're producing in fossil fuel generation. We often do this overnight - during the day is ... a bit more novel!
November 1, 2025 at 12:06 PM
British electricity grid is on a real renewable (windy) roll this week and at the moment, during the day, we're exporting more electricity (to Norway, Ireland, Belgium and Denmark) than we're producing in fossil fuel generation. We often do this overnight - during the day is ... a bit more novel!
British grid had a number of hours last night when exports > fossil fuel generation - ie net zero operation. It will be at least another month or so before the British grid can operate with no fossil fuels due to inertia and other services being supplied in other ways (chart from grid.iamkate.com)
October 5, 2025 at 8:18 AM
British grid had a number of hours last night when exports > fossil fuel generation - ie net zero operation. It will be at least another month or so before the British grid can operate with no fossil fuels due to inertia and other services being supplied in other ways (chart from grid.iamkate.com)
My idle amateur British Energy transition watching happening - last night the the GB Grid was net zero overnight for a long time - got close to renewable energy being 95% of dometic demand, not considering nuclear (I am sure the official statistics will come out a bit later.).
September 7, 2025 at 10:01 AM
My idle amateur British Energy transition watching happening - last night the the GB Grid was net zero overnight for a long time - got close to renewable energy being 95% of dometic demand, not considering nuclear (I am sure the official statistics will come out a bit later.).
Lovely morning swim in Highgate ponds. Not sure how low a water temperature I am going to go to!
September 2, 2025 at 7:14 AM
Lovely morning swim in Highgate ponds. Not sure how low a water temperature I am going to go to!
The (semi) wilds of Northumberland in August - sun burning through hill mist, dew still on the grass, cows and sheep calling out, loud birdsong on top, butterflies labouring between wild flower blooms - a million miles from London and Cambridge
August 24, 2025 at 7:01 AM
The (semi) wilds of Northumberland in August - sun burning through hill mist, dew still on the grass, cows and sheep calling out, loud birdsong on top, butterflies labouring between wild flower blooms - a million miles from London and Cambridge
London apple crop is looking good
August 23, 2025 at 1:50 PM
London apple crop is looking good
Forgot to add photos
July 14, 2025 at 8:02 PM
Forgot to add photos
These volcano plots - x-axis level of skew effect, y-axis adjusted FDR level for significance - orange being significant large effect skews shows we have some strong effects
June 27, 2025 at 2:10 PM
These volcano plots - x-axis level of skew effect, y-axis adjusted FDR level for significance - orange being significant large effect skews shows we have some strong effects
Some - but not all - transposon families have many skewed sites and this is not in a consistent direction in terms of positions in transposons nor transcribed strand or not - but lots of skewing.
June 27, 2025 at 2:10 PM
Some - but not all - transposon families have many skewed sites and this is not in a consistent direction in terms of positions in transposons nor transcribed strand or not - but lots of skewing.
We can zoom in base by base around splice sites here, aggregating across splice sites. Of course some positions wont have CpGs (eg, the AG splice acceptor) but many other sites will have some. We see some strong aggregate asymmetric sites
June 27, 2025 at 2:10 PM
We can zoom in base by base around splice sites here, aggregating across splice sites. Of course some positions wont have CpGs (eg, the AG splice acceptor) but many other sites will have some. We see some strong aggregate asymmetric sites
However, when we split by exon or intron we could see. For me the best analysis was looking in a meta-gene with 3 statistics, each being the asymmetric pair of C/5mC (hemi-methylation), C/5hmC (hemi-hydroxymethylation) and 5hmC/5mC (hemi-hydroxy, hemi-methyl).
June 27, 2025 at 2:10 PM
However, when we split by exon or intron we could see. For me the best analysis was looking in a meta-gene with 3 statistics, each being the asymmetric pair of C/5mC (hemi-methylation), C/5hmC (hemi-hydroxymethylation) and 5hmC/5mC (hemi-hydroxy, hemi-methyl).
One obvious candidate for asymmetry was the action of transcription. When we looked in bulk at sites across genes there was no transcription strand bias to the 5hmC/5mC or other patterns.
June 27, 2025 at 2:10 PM
One obvious candidate for asymmetry was the action of transcription. When we looked in bulk at sites across genes there was no transcription strand bias to the 5hmC/5mC or other patterns.
First up Walter called and isolated only duplex strands keeping track of modifications (now easier in the software updates - a year ago quite a bit more annoying!) and from this we can ask a simple question of distribution of modifications across 9 states (unmodified C, 5mC, 5hmC two strands)
June 27, 2025 at 2:10 PM
First up Walter called and isolated only duplex strands keeping track of modifications (now easier in the software updates - a year ago quite a bit more annoying!) and from this we can ask a simple question of distribution of modifications across 9 states (unmodified C, 5mC, 5hmC two strands)
Flipping this around to variance explained, the amount of variance explained across the 16 loci which was purely additive genetics varied from 15% to 96% of the overall locus (each locus explaining between 1 to 0.1% overall - but the environment explains alot.
June 6, 2025 at 3:12 PM
Flipping this around to variance explained, the amount of variance explained across the 16 loci which was purely additive genetics varied from 15% to 96% of the overall locus (each locus explaining between 1 to 0.1% overall - but the environment explains alot.
Just a reminder that out of the 16 loci we could robustly discover (for a physiological phenotype - heart rate - from a wild-derived population - with a ecologically relevant variation of environment) - 9 of the 16 loci had some sort of interaction term (GxG, GxE or dominance)
June 6, 2025 at 3:12 PM
Just a reminder that out of the 16 loci we could robustly discover (for a physiological phenotype - heart rate - from a wild-derived population - with a ecologically relevant variation of environment) - 9 of the 16 loci had some sort of interaction term (GxG, GxE or dominance)
We had previously noted a striking change in UK<=>Danish genetic exchange from this South Denmark centric to East Midlands/Norfolk to cities - most obviously London/Copenhagen, but also Aaberna and Odense. This now is post 13th cent, most likely 1600+ but could be any (or all) time 1600 to 1850
June 4, 2025 at 10:18 PM
We had previously noted a striking change in UK<=>Danish genetic exchange from this South Denmark centric to East Midlands/Norfolk to cities - most obviously London/Copenhagen, but also Aaberna and Odense. This now is post 13th cent, most likely 1600+ but could be any (or all) time 1600 to 1850
We also now can see a North Norfolk (King's Lynn area) to Southern Denmark, compatible with the documented middle ages trade routes at slightly longer haplotypes (so somewhat later)
June 4, 2025 at 10:18 PM
We also now can see a North Norfolk (King's Lynn area) to Southern Denmark, compatible with the documented middle ages trade routes at slightly longer haplotypes (so somewhat later)
The timing shift means we really can't easily distinguish Anglo-Saxon migrations from the only slightly later Danish (which is really South Denmark) invasions - but what we can say is that the extant short haplotype sharing UK<=>Denmark has a hotspot in Derbyshire and Lincolnshire
June 4, 2025 at 10:18 PM
The timing shift means we really can't easily distinguish Anglo-Saxon migrations from the only slightly later Danish (which is really South Denmark) invasions - but what we can say is that the extant short haplotype sharing UK<=>Denmark has a hotspot in Derbyshire and Lincolnshire
We now have lovely maximum posterior estimates with 50% credible intervals. It is hard - and our theoretical frameworks are not up to it - to handle time, recombination and migration over time and space all together. We've done estimation based on single population model
June 4, 2025 at 10:18 PM
We now have lovely maximum posterior estimates with 50% credible intervals. It is hard - and our theoretical frameworks are not up to it - to handle time, recombination and migration over time and space all together. We've done estimation based on single population model
We also in our simulations explored how big samples needed to be a human context to by able to see GxE effects or dominance effects of the scale we see in this medaka fish study. This is a busy chart - each plot is sample size (y axis, log) needed to get a significant term;
May 29, 2025 at 12:44 PM
We also in our simulations explored how big samples needed to be a human context to by able to see GxE effects or dominance effects of the scale we see in this medaka fish study. This is a busy chart - each plot is sample size (y axis, log) needed to get a significant term;
UK grid in stupidly good excess renewables (sunny and windy!) with low demand - its been net zero (considering export) a couple of times and low price.
May 25, 2025 at 2:48 PM
UK grid in stupidly good excess renewables (sunny and windy!) with low demand - its been net zero (considering export) a couple of times and low price.
Good morning London (and @nanoporetech.com )
May 23, 2025 at 7:11 AM
Good morning London (and @nanoporetech.com )