Opens profile photo
Follow
Moritz Gerstung
@MoritzGerstung
Division head
@dkfz
, professor
@UniHeidelberg
, visiting group leader
@EMBLEBI
. Previously
@sangerinstitute
and
@ETH_en
. We study how cancers (and 🦠) evolve.
Heidelberg, Germanydkfz.de/en/AI-in-oncol…Joined March 2015
806 Following
9,121 Followers

Moritz Gerstung’s Tweets

Pinned Tweet
Moritz Gerstung
@MoritzGerstung
This 👇👇👇 is one of the most exciting pieces of research my lab was involved in during the past few years — and a step change for how we study the growth and evolution of human cancers. Great work by the team!
Quote Tweet
Artem Lomakin
@LomakinAI
From sequencing and evolutionary theory we know that tumours grow as mosaics of clones - but so far we haven’t been able to see them. The labs of @Yates_lab, @MatsNilssonLab and @MoritzGerstung created a new method to map the spatial evolution of cancers: nature.com/articles/s4158
Show this thread
Image
8
163
Show this thread
Moritz Gerstung
@MoritzGerstung
Whatever the reason of frequency dependent fitness, it may be a useful rule of thumb to assume that the daily growth advantage of emerging lineages is likely to drop by 0.02-0.04 in addition to what is expected based on lineage competition.
2
61
Show this thread
Moritz Gerstung
@MoritzGerstung
But this theory is also not without flaws as the lineage zoo exhibited a high extent of convergence — that is independent acquisition of the same mutations. So it may be a bit of both. Would be great to see further in depth analyses and alternative hypotheses.
4
77
Show this thread
Moritz Gerstung
@MoritzGerstung
If the hypothesis were true it could also change the interpretation of why there has been a zoo of coexisting variants around in the past 6 months with slow turnover. Each may have thrived in (and filled) a particular gap in the population immunity.
2
97
Show this thread
Moritz Gerstung
@MoritzGerstung
Omicron and its subvariants descendants can break through our immunity walls. One hypothesis is that susceptibility in the population is heterogeneous and variant specific. So each variant has its own niche in the population, but once it is exploited its advantage drops a bit.
4
151
Show this thread
Moritz Gerstung
@MoritzGerstung
The reasons for the slowing of Omicron variants are not completely understood. For BA.1 it was suspected to be related to its shorter generation time. But the fact that it has been replicated ever since suggests that other mechanisms are at work.
medrxiv.org
Estimation of the test to test distribution as a proxy for generation interval distribution for the...
Background Early estimates from South Africa indicated that the Omicron COVID-19 variant may be both more transmissible and have greater immune escape than the previously dominant Delta variant. The...
1
68
Show this thread
Moritz Gerstung
@MoritzGerstung
It was quite different for Delta and Alpha which spread at remarkably constant rate. But times and immunity were very different then.
Quote Tweet
Alex Selby
@alexselby1770
About the growth advantage of the Delta (first found in India) variant over Alpha (first found in the UK). Lots has been written about this with many different opinions, but let's take a look at this graph: (thread: 1/n)
Show this thread
Image
1
66
Show this thread
Moritz Gerstung
@MoritzGerstung
For XBB.1.5 only few countries have sufficient data to calculate a long term growth advantage, so it seems constant above. But the patterns in the US and U.K. are suggesting it has also come down more quickly than expected. I’d expect the drop to become more evident.
Image
2
75
Show this thread
Moritz Gerstung
@MoritzGerstung
In fact this slowing was observed for almost every Omicron lineage. There is large variation between countries though, in part because of the low numbers at low incidence. But the trends are clear that the initial growth rates dropped down on average between 0.02 to 0.04.
Image
Image
3
69
Show this thread
Moritz Gerstung
@MoritzGerstung
In slowing down XBB.1.5 follows a pattern that has been noted also for BQ.1.1 or XBB.1.1. Their initial fitness (daily increase of variant share) was higher than their long term advantage in a multi-lineage model with constant differences between variants (coloured lines).
Image
Image
2
77
Show this thread
Moritz Gerstung
@MoritzGerstung
There are some signs now that XBB.1.5 is loosing steam as it spreads through the wider population. The share of cases has increased more slowly in the US and UK, recently. A speculative thread why this might be.
Image
Image
22
484
Show this thread
Moritz Gerstung
@MoritzGerstung
An updated global analysis of XBB.1.5 * XBB.1.5 has spread quickly across the globe reaching shares between 1-10% (nowcast) * XBB.1.5's share doubled every 7-14d * The reported share in the US has stagnated below 25% (but little data since Xmas - may rise with data coming in)
Image
Image
Image
7
246
Show this thread
Moritz Gerstung Retweeted
Jakob Nikolas Kather
@jnkath
Great start for 2023! Our new paper is out in
@NatureMedicine
, led by
@twittlabs
from
@uni_mainz
. We use a multi-modal deep learning pipeline to predict outcomes in #colorectal #cancer.
@EKFZdigital
@NCT_UCC_DD
@NCT_HD
. Shout out to
@AI4Pathology
who pioneered this field.
Quote Tweet
Sebastian Foersch
@twittlabs
Happy to share our latest publication in @NatureMedicine 🩺🧬🔬 nature.com/articles/s4159. We designed a multistain deep learning model and trained, validated, and tested it on imaging data of different immune cell subtypes of over 1000 patients with #colorectal #cancer. Our …
Show this thread
Image
Image
2
108
Moritz Gerstung Retweeted
Bloom Lab
@jbloom_lab
I want to discuss relationship between ACE2 affinity and fitness (transmissibility) of human SARS-CoV-2 variants. Topic is of interest because defining feature of the new XBB.1.5 variant is increased ACE2 affinity relative to its parents:
Quote Tweet
Yunlong Richard Cao
@yunlong_cao
The superior growth advantage of XBB.1.5 has been well-documented by many colleagues @JPWeiland @LongDesertTrain @EricTopol. Here I'll add some experimental data: 1) XBB.1.5 is equally immune evasive as XBB.1, but 2) XBB.1.5 has a much higher hACE2 binding affinity. 1/
Show this thread
Image
14
401
Show this thread
Moritz Gerstung
@MoritzGerstung
Further details on XBB.1.5 can also be found in this instructive thread.
Quote Tweet
T. Ryan Gregory
@TRyanGregory
I think XBB.1.5 (Kraken) is very instructive about how dynamic and wide-reaching SARS-CoV-2 variant evolution is, and why confident dismissals of each variant's risk and a focus only on snapshots at one time in one place are foolish. 🧵
Show this thread
1
63
Show this thread
Moritz Gerstung
@MoritzGerstung
And it is also worth noting that variant trackers such as
@siamosolocani
have pointed out the rise of S:486P carrying lineages more than a month ago.
Quote Tweet
Federico Gueli
@siamosolocani
Replying to @EllingUlrich @EricTopol and 8 others
Her ei want to highlight also an uptrend i am seeing quite clearly for all "S:F486P carrying lineages" (XBF,XBB.1.5,CJ.1,XAY,XBC) @MoritzGerstung @TWenseleers @PeacockFlu @CorneliusRoemer @K_G_Andersen cov-spectrum.org/explore/World/
2
79
Show this thread
Moritz Gerstung
@MoritzGerstung
A yardstick to judge the pandemic potential is the fitness relative to the variant mix at emergence. XBB.1.5's fitness is comparable to that of BQ.1(.1) and XBB(.1). It is lower than that of BA.1/2 and BA.5, though, which rapidly replaced previous variants and caused surges.
Image
2
108
Show this thread
Moritz Gerstung
@MoritzGerstung
XBB.1.5's growth advantage is thought to stem from its S:486P mutation. Unlike XBB's S:486S mutation, it retains ACE2 binding affinity while still evading antibody neutralisation. It took long to arise because it requires two nucleotide changes.
Quote Tweet
Bloom Lab
@jbloom_lab
A scientifically interesting aspect of XBB.1.5 is we pretty much understand what mutation made it so transmissible, the mechanism by which the mutation acts, and why it took so long for the mutation to emerge.
Show this thread
1
103
Show this thread
Moritz Gerstung
@MoritzGerstung
It's not uncommon, though, for a variant's spread to slow slightly after a steep initial rise. For example BQ.1.1 arrived a month or two later than initially projected. Similar effects were seen for BA.2.75 and even BA.1/2 and may be caused by heterogeneous susceptibility.
1
79
Show this thread
Moritz Gerstung
@MoritzGerstung
XBB.1.5's share is rising globally, too. However, it spreads slightly slower than in the US with relative doubling times between 8-15 days. This makes XBB.1.5 currently the fastest spreading lineage, followed by CH.1.1. It could possibly replace BQ.1.1.
Image
1
104
Show this thread
Moritz Gerstung
@MoritzGerstung
Last SARS-CoV-2 variant update of 2022: * A three-way stalemate between BQ.1*, XBB* and BA.2.75* variants. * BQ.1* dominant in Western, XBB* in Eastern hemisphere. * Proportions change slowly as variant fitness equalise out and other BA.5* variants have largely disappeared.
Image
Image
Image
Image
6
298
Show this thread
Moritz Gerstung Retweeted
Yunlong Richard Cao
@yunlong_cao
Our paper regarding Omicron convergent evolution is out on
@Nature
. In this story, we analyzed the immune evasion capability of ~50 convergent variants and explained how RBD mutations suddenly emerged convergently due to a more focused immune pressure.
nature.com
Imprinted SARS-CoV-2 humoral immunity induces convergent Omicron RBD evolution
Nature - Imprinted SARS-CoV-2 humoral immunity induces convergent Omicron RBD evolution
36
1,089
Show this thread
Moritz Gerstung Retweeted
Artem Lomakin
@LomakinAI
Our review is out! In it, we explore: 🏙️ How tissue architecture influences cancer evolution 🏞️ What do we know about how cancer clones co-evolve with their ecological niches 🔬 The latest tools to comprehensively describe spatial biology of cancer evolution
Quote Tweet
Moritz Gerstung
@MoritzGerstung
Interested in spatial biology and cancer evolution? New spatial genomics technologies start unraveling how somatic evolution, tissue microanatomy and tumour microenvironments interact. New review with @Yates_lab @LomakinAI and @zaira_sef nature.com/articles/s4157
Show this thread
1
28