YT YouTube2Text

YouTube2Text → Transcripts → Sarah Teichmann

Sarah Teichmann — Transcript

by Fragile Nucleosome · 4,686 words · 765 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00my um great pleasure to introduce our
0:02second speaker um Dr. Sarah um Techman.
0:06So Sarah did her PhD at the MRC
0:09laboratory of molecular biology in
0:11Cambridge, UK and he uh she was a belt
0:15memorial fellow at University College
0:17London. In 2001, she started her own
0:20research group in MRC laboratory of
0:23molecular biology and moved to the
0:25Walcon genome campus in 2013.
0:29And in 2016, she was appointed head of
0:32the cellular genetics program, Ali
0:34Sanger. And in 2024, she then took
0:38another leadership role as a chair in
0:40Stanell medicine at the University of
0:42Cambridge. She's co-founder and also
0:44co-leader of the international human
0:47cell atlas consortium which aims to
0:49create reference maps for cells across
0:52all human tissues and has grown to
0:54include over 3,000 members across the
0:56world. Her laboratory develops as well
0:59as um applies cell atlas techniques to
1:02understand human tissue architecture
1:05with a particular focus on how cellular
1:07diversity is generated in the immune
1:10system and also throughout development.
1:12Her work has been recognized by numerous
1:15award including the amble gold medal
1:17genetic society Mary lines award among
1:20many many other awards and today um it's
1:24our great pleasure to have her
1:26presenting about multiomic Atlas of
1:28human skeleton development. Welcome
1:31Sarah. Thank you so much Sky for that
1:33kind introduction and it's a pleasure to
1:36to present this work um which we
1:38published at the end of last year um and
1:42it's slightly different from the title
1:44that I indicated but I want to kind of
1:46dig into this work because it had a um a
1:50a a component of of the epiggenome
1:54which we leveraged for gene regulatory
1:56network inference and genetics. I want
1:58to just sort of give a a very quick
2:00introduction into the the technologies
2:02that we use in the human cellless
2:04community and in projects like this and
2:06they're obviously single cell and
2:08spatial genomics and you know you're all
2:10aware of this resolution revolution that
2:12we've lived through for the past 15 plus
2:15years um where we are we are now able to
2:20map the nucleic acid content and um
2:24multimodal
2:26um features of individual cells and
2:29nuclei. And of course, combining that
2:32with spatial transcrytoics, with highly
2:35multiplex spatial mapping methods and
2:36tissue sections is a really powerful way
2:40to map human cells. And I'm really
2:42delighted that Ishtak who who preceded
2:45this talk kind of was also talking about
2:47human developmental samples, which is
2:49exactly what I'll be speaking about too.
2:52Um, and and these methods really sort of
2:54open up our ability to understand the
2:57molecular and cellular composition of
3:00human tissues in ways that weren't
3:03possible previously um because of the
3:06the limitations of of um you know
3:09modulating genetics and and manipulating
3:12human tissues. And it's really these
3:14technologies together with the data
3:16science that stitches them together that
3:18buoyed the launch of the human cell
3:20atlas almost 10 years ago now. And of
3:22course, this is
3:24our community grassrootsled project um
3:28that that that's an open project where
3:30we invite any anyone interested in this
3:33this this endeavor to join and the link
3:36is down at the bottom where our aim is
3:38to create a reference map of ourselves
3:42kind of from a basic understanding
3:44and discovery science point of view. And
3:48of course studying the human there's
3:50also immediately a diagnostics and and
3:53treatment implication and sort of
3:56translational relevance and and just to
3:59sort of um say this is a slide that's
4:01out of date but we're now at over 4,000
4:04members from over 100 countries around
4:06the world and you see sort of Europe and
4:09and North America and also um Latin
4:12America, Africa, Asia and so on has
4:15increasing kind of members and repres
4:17representation that uh contributing to
4:20the project. [snorts] The phase that
4:22we're in right now is really important
4:24and I want to highlight this this draft
4:27atlas assembly phase. And what this
4:30means is that we are integrating data
4:32sets and releasing reference data
4:34objects at data.humansellatlas.org
4:36the human cellless data portal. And so
4:39these are uh objects that are put
4:41together by the 12 postocs that are
4:43funded by the trans Zuckerberg
4:44initiative to integrate these data sets
4:47and work with the bio networks. We have
4:5018 biological networks that have PIs
4:53that are experts in the different organs
4:55and systems and curate do a careful
4:57curation of cell states. And so I want
5:00to emphasize that and encourage people
5:01to go and look there are now more
5:04approved atlases that we released. And
5:06this morning we approved the pancreas
5:07and the uh the the oral cranial facial
5:11atlas, the m the oral cavity basically.
5:14Uh and so that you know this is a really
5:15exciting time. We're obviously kind of
5:18still actively extending the human cell.
5:20This is far from complete with
5:22multimodal data with spatial data with
5:24more tissues pushing towards a more
5:27comprehensive mapping. We we we do have
5:31um on the data portal over 100 million
5:33cells from the 18 different biological
5:36networks and you can see the organs and
5:37systems sort of indicated around here.
5:40My story today is going to be related to
5:43muscularkeeletal
5:45network as well as
5:49human development embryionic [snorts]
5:51and and fetal development kind of along
5:53the lines that that Ishtak mentioned.
5:56And and so what I'll be what I'll be
5:58presenting is um human skeletal
6:01development where we set out to
6:03characterize osteogenesis and
6:05condrogenesis and gain insights into
6:07bone and joint conditions. And we have
6:08already worked on limb development
6:11previously and published um kind of
6:15insights into
6:17the the the condraittes. So the
6:18cartilage um that makes our tendons and
6:21and ligaments as well as osteoccytes
6:24that contribute to bone. And in terms of
6:27um oh just a sec
6:31need to change the pointer to to show
6:35these videos. So in in first trimester
6:40um embryionic osteogenesis the the bones
6:43and cartilage develop. And here what we
6:46have is a light sheet imaging from um
6:49Rafa Blan Alan Shadow's lab in Paris
6:52marking a collagen that's a marker of
6:55condrogenesis and a transcription factor
6:57SB7 that's a marker of osteoggenesis.
6:59So these are two lineages cartilage and
7:02bone that develop from the same meenal
7:05progenitors and there are two different
7:07processes um uh intra intromemebranous
7:10oification that makes our skull cap the
7:13the the top of our head and um
7:17endocchondrial oification that
7:18contributes to the other bones in the
7:20body and the the skull base. And you can
7:23see this sort of development kind of in
7:26a
7:28in this translucent um you know
7:30stunningly beautiful image of of the
7:32embryo here where [snorts]
7:34Raphael and allow then zooming in
7:36showing the skull cap has a different
7:38developmental origin to the skull base.
7:41[snorts] So, as I said, the SP7
7:44transcription factor is marking the
7:45skull cap and the
7:47specific collagen is marking the skull
7:49base because they're they're obviously a
7:51single sort of anatomical structure
7:53together, but have different origins and
7:56different
7:58mechanisms for making the bones. And if
8:00we look a couple of weeks later in in in
8:02pregnancy and in human development, we
8:04see these [snorts]
8:07um kind of more mature um
8:12um ex more extensive
8:15uh coverage of of the skull cap and the
8:17skull base through these um these two
8:21different processes.
8:22Intramebran intramebranous ocification
8:25and endocchondrial ocification.
8:29um in in the skull cap the introme
8:32intramebranous ocification is the main
8:34mechanism. So here again just to um
8:37emphasize that
8:40the the different um plates effectively
8:44that that um contribute to our skull are
8:49uh anterior regions of the skull cap and
8:51skull base. And you have these uh
8:53sutures and um [snorts] um the coronal
8:58suture and sagittal sutures that uh are
9:01at the the interfaces between the the
9:03the the plates. Um there's ocification
9:07across regions during development and
9:09the the skull cap is formed by
9:12intramebranous ocification and Ken toe
9:15who was an incredibly talented um
9:18clinical PhD student traininee in the
9:20lab and has now gone back to surgical
9:23training dissected these regions
9:25extremely carefully for then paired
9:28single nuclear and single single nuclear
9:31RNA and attack sequencing. So multi ohm
9:33analysis of the nuclei which you can
9:37it's the tissue soft enough that you can
9:38extract it basically from this first
9:40trimester samples and um and then also
9:45analyze uh the sections by vizium
9:49spatial transcrytoics and this is shown
9:51here um [snorts] uh sort of indicating
9:55the skull the the skull cap the skull
9:57base and then also the appendicular
9:59skeleton joints in the shoulder the hip
10:01and the knee
10:03um that were dissected at different time
10:05points in in first trimester development
10:08year 5 6 7 8 9 10 11 12 postconception
10:11weeks and then analyzed by um
10:15multi ohm so RNA plus attack sequencing
10:18at these different time points for the
10:19different um bone and joints and then um
10:23also by vizium and incite 155th fiveplex
10:27incitu sequencing which we carried out
10:28with with Kenny White um in in Omar
10:32Batra's lab um and the the the process
10:38of um intramebranous oification sort of
10:42takes place throughout where the onset
10:45of oification is kicked off in the skull
10:47cap throughout this period. Whereas in
10:50the other um bone and and joint regions
10:54there's this misenal condensation
10:56interzone formation and then formation
10:59of of cartilage primordia and
11:01paricchondrium and and then finally
11:04oification through endocchondrial
11:06oification which is sort of a different
11:08developmental process.
11:11If we kind of look at the the cell
11:13states of these 300,000 um nuclei that
11:17we profiled by Multium, the meseno
11:20states are shown on the right hand side
11:21and you can see a lot of different um uh
11:24progenitor and um
11:28developmental states kind of um all
11:32projected together from from the
11:34different regions. Over on the right
11:35hand side here, um
11:39you've got condondraite progenitors of
11:41different at different stages. Um
11:44hypertrophic condraittes
11:46uh and and articular condondraittes. If
11:49we focus on the the the the the skull
11:52cap and craniogenesis,
11:54we describe these previously unreported
11:56human cell states of the cranium misim,
11:59the suture misenheime, and then a
12:02pre-ostoblast progenitor prior to the
12:04osteoblast and osteoccy differentiation.
12:06And because we have um [snorts] we we
12:10can sort of plot this in in this pseudo
12:13time um developmental cell cell pseudo
12:17time here um and and and plot the
12:20markers that are differentially the
12:22genes that are differentially expressed
12:24along this this trajectory.
12:27And then what I'd like to um you know
12:29what's really beautiful is that this
12:31also maps kind of onto onto um
12:37different locations in in space in the
12:40skull cap. And just to orient ourselves
12:43this is 50 micron vizium spatial
12:45transcripttoics where each voxil you
12:47know is a group of cells. We can
12:49deconvolute the cells with
12:50celltolocation which is a probabistic
12:53model um that gives us the individual
12:56states in the voxels and um the suture
12:59the coronal
13:01sutures basically along along in this
13:04region you've got der the dermal layer
13:06the skin on the top and then mining
13:10um along the bottom then then the brain
13:12basically below that and
13:16um in fact the the different uh the the
13:20different cell types that I just
13:21mentioned are organized in um in a
13:26zonated way basically along the skull
13:28cap and we can use our organ access um
13:32modeling framework that's on GitHub that
13:35we published as part of a a whole thymus
13:38cell atlas model um in order to
13:40calculate kind of the different uh the
13:43different zones in a progressive way
13:46that represent the different cell uh
13:48cell states or the different sort of
13:51tissue niches if you like kind of um
13:55along the um along the skull cap. So
13:58you've got suture zones and then
14:01different osteogenic zones um sort of in
14:04a progressive way and the the the the
14:08[snorts] osteoccytes are are basically
14:10enriched in the um
14:13in the osteogenic zones and and the
14:15progenitors basically earlier on
14:19um because this is multi ohm data so we
14:22don't have the the the beautiful histone
14:23modifications that you saw in Ishtiaak's
14:25talk what we have is is open versus
14:27closed chromatin And so we can define
14:30um combinations of of enhancers and and
14:33and actively transcribed genes using the
14:36scenic plus um gene reguy network
14:39inference pipeline and and [snorts]
14:41calculate these enhancer
14:44gene transcription factor active gene
14:47sort of modules and that's what you see
14:49over here on the right hand side where
14:51we quantify the activity of these
14:52different transcription factors in the
14:55different um in [snorts] the different
14:57cell compon compartments, the different
14:58cell lineages. And you so you can see a
15:01set of transcription factors that's um
15:05um sort of more active in the in the
15:08earlier cell compartments and and and
15:11then this SP7 transcription factor that
15:13I meant that I mentioned marking the
15:16intromebrinous oification and the
15:18osteoblast and osteoccytes is coming up
15:20here in these more mature compartments
15:23and we can you know represent these
15:25regulons
15:26with their their dominant transcription
15:29factor. factor basically in and and
15:32correlate different modules kind of
15:34together within cells. And so there's
15:37sort of [snorts] modules that are
15:39essentially activating or pushing
15:41forwards osteogenesis versus modules
15:44that are kind of inhibiting osteogenesis
15:48if you like and and they're shown here
15:49in the red and the blue.
15:52Um so so using that sort of um network
15:56approach
15:58um we we can
16:01essentially [snorts]
16:03um calculate a a hierarchy of
16:05transcription factors that are
16:07maintaining the progenitor pool and um
16:10versus ones that are activating
16:13um activating uh the osteogenic
16:17phenotype essentially in in both
16:20intramebranous and in um um
16:24endocchondrial
16:25uh ocification.
16:30and and and then if we if we again sort
16:33of zoom in and map uh the locations of
16:36the transcription factors and their
16:38activity onto the the skull cap in space
16:41in the spatial transcripttoics, what we
16:44can see is the um
16:47um transcription factors that are kind
16:50of poisoning for osteogenesis in the
16:52sutra misanky versus transcription
16:55factors that are that are um maintaining
16:57patency like twist one and lmx1b.
17:00And then also um you know correlate how
17:05these transcription factors how these
17:07key transcription factors link to
17:09genetic conditions like cranioinostosis
17:12where you have fusion of the sutures.
17:15And so this um you know twist one is
17:18expressed in this region maintaining the
17:21the the separation of the the the skull
17:25plates. Um and if there if that's
17:28mutated then then there's fusion
17:30cranioinistosis of the the skull plates
17:32which is a bad thing. Um in similarly uh
17:36when we look at um
17:39[snorts] at at other um other other
17:43signaling or factors or receptors or
17:46transcription factors involved in in for
17:48instance um digit you know the the
17:51development of the bones and the digits.
17:53We can plot basically where these genes
17:57are active or penetrating during first
18:00trimester development in the the not
18:03only the skull cap but also the other um
18:06you know the other parts of other bones
18:08for instance in digit development and
18:09relate essentially the the the phenotype
18:13from proximal digit only distal digit
18:15only and so on to the the location where
18:18the genes are acting during development.
18:21And so this is um you know it's it's
18:24giving us a very high resolution fine
18:26grained insight into the
18:29>> [snorts]
18:29>> um the mechanisms of these genetic
18:31conditions
18:33and you know including the sort of
18:35cranioacial cleft which is enriched in
18:37cranial bone
18:39pytorranial disostois enriched in
18:42cranial bone and limb paricchondrium and
18:44so on.
18:46um we collaborated with the Zen group in
18:48also linking um expression of of G-W was
18:53fine mapping of of G-W was signals so
18:56where you've got a genetic association
18:58with with adult diseases such as
19:00osteoarthritis and rheumatoid arthritis
19:04into the the developmental cell states
19:06and you know we were discussing this in
19:08my lab this morning uh whether this
19:10makes sense and you know I would I would
19:13argue there are two there are two well
19:16not only did the the reviewers kind of
19:19accept this but you know beyond that I
19:22would argue that it it is likely and
19:26there is evidence that that some of the
19:28conditions that we experience as adults
19:30and that are late onset are actually
19:32sort of already poised um in in terms of
19:37uh you know subtle variations during
19:39pregnancy in in our embryionic and fetal
19:41developmental stages and and So, you
19:44know, understanding the genetic variants
19:47that increase our risk of various adult
19:50diseases during development,
19:53I would argue is is relevant and is of
19:56interest. And obviously, you know, if we
19:58find map variants and then understand
20:00how they're modulating binding of
20:03transcription factors and influencing
20:05changes in gene expression in cells
20:07during development, that can give us
20:10mechanistic insights. And so this
20:12exercise, you know, this is this is one
20:14example
20:15um uh where we're relating adult risk
20:19genetic risk factors that are mostly
20:21non-coding by the way into um you know
20:24they're they're acting through distal
20:26enhancers and repressors and so on into
20:29um the the developmental data is is is
20:32is kind of you know one example. And um
20:37the way we went about this was using a
20:40new method called snip to cell which we
20:41published in this in this paper um that
20:44was led by Ken toe with Lee Jang Feay
20:46and Patrick Pet uh who who worked
20:49together to develop this method. And the
20:51the the principle is that you take the
20:53G-W was summary statistics, map them
20:55into the um the multi ohm data uh into
20:58the the enhancers and the gene regulator
21:00network and look for enrichment of
21:03modules of of genetic variance in cell
21:06types through as they sort of percolate
21:10through the gene reguary network that's
21:12calculated from um from the scenic plus
21:16data. And so you can kind of propagate
21:18variance through uh modules and
21:20calculate subn networks that are that
21:22are affected. And and indeed when we do
21:25this this this when we use this snip to
21:29cell pipeline that uses fine mapping of
21:31GW was and FGW was uh package with the
21:35single single nuclear multiom data. We
21:37can identify enrichment of osteogenic
21:40cell types uh in in HIP.
21:44um but um
21:47condondrogenic cell types in in knee
21:51osteoarthritis. And so what it's what
21:53it's suggesting is that knee
21:54osteoarthritis is more um you know
21:57associated with cartilage and and cart
22:00and and sort of poised through cartilage
22:02development and hip osteoarthritis is
22:04more connected to osteogenic bone cell
22:07types in in in the hip. [snorts] Um
22:14so basically that sort of takes me to
22:16the end of this story and um just to
22:20summarize what we've um discovered are
22:23are cell types and tissue niches in in
22:26skeletal development that were
22:28previously unappreciated
22:30in in in mammals. So not just in humans.
22:34Um
22:36that uh uh uh also we use scenic plus to
22:41uh identify gene regu network triads
22:44using the multiome data and and define
22:46transcription factors, regulatory
22:48regions and genes that are linked
22:50basically in in in these uh different
22:54cartilage and bone developmental
22:57processes that occur in the different
22:58regions of our skeleton and then connect
23:01to rare disease and common disease
23:03genetics.
23:04and and gain more detailed insights into
23:06the mechanisms of these genetic
23:08conditions.
23:10Basically, by mapping the rare disease
23:12genes in terms of their locations and
23:13where they're penetrating during
23:15pregnancy and then also gaining insight
23:17into where adult onset disease uh risk
23:22factors may be acting during
23:24development.
23:26So I've mentioned Ken who um
23:30worked closely with Lee Shang and
23:32Patrick who are two incredibly talented
23:34um uh computational [snorts]
23:37data science and AI postocs in the
23:39group. Patrick has now gone to uh
23:42relation a biotech company. Our
23:44collaborators um were Mosh Hanifa, Roger
23:47Barker and I've also mentioned Omar
23:49Bactctor already and also Alan Shadowal
23:52for the beautiful light sheet imaging
23:54and Chris Buckley who co-supervised Ken
23:56for his PhD and is an an expert
23:59rheumatologist in Oxford and this is a
24:01picture of our lab and um these are my
24:05disclosures.
24:07Thank you for your attention and I'd
24:08[clears throat] be delighted to take
24:09questions.
24:12Thank you Sarah for the uh wonderful
24:14presentations. Um if you have any
24:16questions you can raise your hand or you
24:18can type your questions um in the Q&A
24:21box so that um Sarah can answer it
24:24directly.
24:29Let me just make sure
24:43So the first question is um
24:47[clears throat and laughter]
24:48[gasps]
24:48um it's a praise.
24:59How easy is it um to collect a lot of
25:02samples for your like this single cell
25:06experiment?
25:14Sorry, I'm just answering the question.
25:16Um so, so sample access is a crucial
25:19point and thank you so much for that
25:21question which I kind of skipped over.
25:24um you know all this data was enabled by
25:26the the generosity of the women who are
25:28donating their tissue um in a a social
25:32termination.
25:33So similar to what Ishtak said and in
25:36the UK we have a wonderful tissue bank
25:38called the human developmental uh
25:40biology repository www.hdbr.org
25:45and um that that repository is based at
25:49um London and Newcastle and provides
25:52tissues uh on a bio bank basis uh to to
25:56scientists in the UK and also uh to
25:59scientists in other countries if you can
26:03uh work with them uh to show that you
26:05have the the appropriate ethics from
26:07your institution the ethical approval
26:10>> um the these samples were also uh
26:14collected under a separate ethics held
26:16by Roger Barker in Cambridge uh he's one
26:19of the PIs that I acknowledge on the
26:21last slide and Roger has run a um a
26:25research program for over 20 years on um
26:29um dopamineergic neurons in the
26:31developing brain in order to uh better
26:34understand and treat Parkinson's disease
26:36and and and [clears throat] we worked
26:38with him. I'm a co-PI on that research
26:41ethics project where we also collect
26:44samples locally from the uh tertiary
26:47teaching hospital in in the Cambridge
26:49University Clinical School.
26:52Nice. I think Jackson has a question
26:54that um he's going to ask himself.
27:00>> Got that closing camera. Yeah. Uh that
27:03was really fascinating, Sarah. And I'm
27:06wondering um how much information you
27:09can kind of glean or infer from these uh
27:12networks about like the
27:16origin of all of these different cell
27:18types in the skeletal system. um like
27:22from like the transcription factors that
27:24you're pulling out or anything else
27:26about the chromatin accessibility, can
27:28you identify subsets of cells that might
27:31be like neural crest derived versus the
27:34cells that come from other ectoermal
27:36lineages? Yeah. So the this is all
27:39mezeno lineage and um sorry let me just
27:43go back.
27:46Um
27:53yeah so these lineages are derived from
27:55different condensates essentially in the
27:59um in the developing embryo that are
28:02meenal
28:03um meenal derived. um there are the the
28:08anterior part so the anterior plates
28:12are are are neural crest derived and
28:16have
28:17um so I didn't I didn't kind of go into
28:21this actually here
28:23um
28:26yeah so so it's a great question so
28:28basically these parts are are neural
28:30crest derived and these parts are mezenl
28:33derived and the neural crest test
28:35derived
28:36um
28:38do have a kind of um
28:42scar or like remnants of transcription
28:45factors that give us the clue that
28:46they're neuro crest arrived in um
28:51and and and
28:52the um the dissection also
28:57um kind of has a you know there the
29:00dissection is is um is kind of
29:04platebased and we can also um trace that
29:08basically from the spatial
29:09transcripttoics
29:12in terms of
29:15non-coding kind of regatory elements
29:19you know and epigenetics there should
29:21also be kind of a scar or like a a
29:24memory but I I can't remember exactly
29:27whether we dug into that but you know
29:29the I I should say that the data you
29:31know all of this data is publicly
29:32available for deeper
29:34or reanalysis,
29:36further analysis, reuse and so on so
29:38forth and um both on the human cell data
29:43portal and also from our own
29:44supplementary website.
29:50>> Yeah, that's thanks so much. Looks
29:52[clears throat] like there is another
29:52question in the Q&A.
29:56Um
29:57>> okay, let's see. Um sorry,
30:01>> two more questions now. Yeah, two more
30:03questions. You suggested that the embryo
30:06marks adult disease. How did you uh in
30:10amber lease mark being activated in the
30:13adult?
30:17Yeah. So, so the way I'm thinking about
30:19this and and this is kind of, you know,
30:22I sort of welcome critical feedback and
30:25input and kind of conversation around
30:27this, but the way I think about this is
30:29that, you know, our DNA is kind of the
30:31blueprint
30:32um of of ourselves. And remember, a lot
30:35of these variants that are risk factors
30:37for adult onset diseases are in
30:39non-coding regions. And so that that
30:42molecular variant [snorts] that you know
30:45is different in different people is
30:47going to have you know a mechanistic
30:50impact on transcription factor binding
30:52affinity on gene expression levels
30:55whether it's in [snorts] you know
30:57whether it's it's it's active I'll say
31:00in in development in in a certain
31:03context and certain cell compartment
31:07equally as in adult if it becomes comes
31:11sort [snorts] of unleashed if you like
31:12in disease
31:15and but the the specific
31:18like the the the the specific cell type
31:20or location in the body may not be
31:22identical.
31:25So
31:27my feeling is basically that that it's
31:29you know from a it's worth tracing the
31:33impact of these variants that we know
31:35from G-W was associations
31:38are risk factors for disease. It's worth
31:40understanding also how they are
31:42influencing you know or whether they're
31:45they're
31:47they're active
31:50in influencing subtle changes in in
31:52human development.
31:54So that that's how I think about it. And
31:56I don't know if that makes sense to
31:58everyone, but I'
32:00um [snorts] you know that's that's
32:02basically
32:04um why I think it's useful to actually
32:07do this this analysis.
32:14>> The second question is how about
32:16sinosis? Anything special with regards
32:19to the rest of the skull?
32:22>> Sinuses. Okay.
32:26Um,
32:28yeah, they cause a lot of pain kind of
32:30when you've got a an infection. Um,
32:35so I'm not sure when
32:39Yeah, I'm just going back to the the
32:42video. Um,
32:46so we didn't
32:49dissect the sinuses specifically. So we
32:51didn't sort of we didn't although you
32:54can see the the the regions that we that
32:57we focused on as I said were sort of
32:59skull cup skull base and then you know
33:01the knee
33:03um [snorts]
33:05um um the the shoulder the hip and the
33:07knee
33:10the dissections what what I should
33:12emphasize is that the surge the surgical
33:15trainee who did this can tow you know
33:18this is really like micro micro surgery
33:21and dissection or dissection.
33:25I think it would be super interesting to
33:27look at the different the you know the
33:29other the sort of oral cranial facial
33:33anatomical components
33:36um and and and maybe the way to tackle
33:39that is not by dissection but by you
33:42know getting the right sections for
33:44spatial transcrytoics
33:47and and sort of serial sectioning to
33:49kind of reconstruct in three dimensions.
33:52Um,
33:55so I think, you know, things like the
33:56sinuses, the inner ear,
33:59even the dental kind of development are
34:02super interesting, but we didn't we
34:04didn't tackle that part kind of in this
34:06study, but [snorts] it would be it would
34:08be a really interesting project for the
34:10future.
34:14>> Right, I think um that's pretty much all
34:17the questions we have for your talk and
34:18it's a lot of questions and thank you
34:20for answering. Um we are actually right
34:22on time. Thank you again um for both
34:24speakers for the wonderful presentations
34:26and also for everyone for participating.
34:29Yeah, with that we are going to end this
34:31um seminar.

About this transcript

This page contains the full transcript of Sarah Teichmann by Fragile Nucleosome, generated from the public captions YouTube serves with the video. The transcript has 4,686 words across 765 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.