3/23/22 – Dr. Jeffrey Gruen –The New Yale Program for Learning Disabilities Research

March 24, 2022

Information

For CME Credit, please read the CME announcement for this lecture.

For Community Practitioners, please read the following CME announcement.

ID7599

To CiteDCA Citation Guide

00:00Spring.
00:05We have a few announcements,
00:06upcoming grand rounds next week,
00:09March 30th is one of our.
00:11A special lecturers named lecture
00:14and part of our Centennial series,
00:17the C DAF Cook Lectureship.
00:19Those of you who don't remember
00:21that Cook was a former chair of
00:24the Department of Pediatrics.
00:27And Kinari Webb,
00:28who's in family medicine and
00:29founder of health and harmony,
00:31will be speaking on healing
00:33the planet by recognizing the
00:36fundamental sickness of colonialism.
00:38Uh.
00:40Then on April 6th we're going
00:42to be having care rounds,
00:43which are conversations and
00:46reflections experiences in the
00:48Children's Hospital and look forward
00:51to that interactive grand rounds.
00:57Another announcement coming up on
00:59this Monday or monthly pediatric
01:02educational learning community
01:03series will be led by Galloping
01:06Me Soma and Gunjan Tiagra,
01:09both from the section of emergency
01:11medicine on introduction to
01:12qualitative research methods,
01:14exploring factors that influence
01:16the choice of academic pediatricians
01:18by underrepresented minorities.
01:25And also Joe reminder.
01:27Another one of our major
01:29events for a Centennial.
01:30Is the Richard Ehrenkrantz Neonatology
01:33Symposium and this year's topic
01:35is ethical issues in neonatology?
01:37This will be taking place virtually Tuesday,
01:40April 5th from 1:00 to 5:30.
01:44There will be a incredible
01:47lineup of speakers and talking
01:50about these important topics in
01:52ethics related to neonatology,
01:54so please join us.
01:59There are no conflicts and no contact,
02:02no disclosures that need to be made.
02:06This is accredited for see me
02:09by the Yale School of Medicine.
02:11There's the number.
02:12The text to the CME office
02:15and will also put this in the
02:18zoom chat at various intervals
02:20throughout the grand rounds.
02:25And with that I'm going to
02:28turn it over to Mark Mercurio,
02:30to introduce today's speaker.
02:38Thank you Cliff.
02:39It's an honor to introduce my
02:41friend and colleague Jeff Grouin.
02:42Today I'm gonna just share my
02:44screen for a moment if I could.
02:49Let's go with this and share and
02:52are you looking at my slides now?
02:54So this this of course is doctor Jeff Broin
02:57who is known to you all as a professor
02:59of Pediatrics and genetics here at Yale.
03:01He's certainly one of the pillars of of
03:05Yale and neonatology, Yale Pediatrics,
03:08and Yale Science, but of course he
03:10didn't start out as such a big cheese.
03:13He was got his BS and MD at Tulane,
03:16came to Yale where he did his residency
03:18and fellowship and Jeff and I traveled
03:20this journey together ever since.
03:22And so it's been my absolute pleasure
03:24to watch him rise through the ranks.
03:26So let's see if we can get this slide
03:29to advance and apparently we can't.
03:31Let's try that.
03:32So here we go back a little bit
03:34in time and we see this here.
03:35I don't know if you can see my cursor.
03:36We see the second from the left.
03:37Standing up is a young Jeff grew in with his
03:40dog with his wife Susan here and Michelle.
03:42I see everybody.
03:43We had a lot of babies were
03:44born in Pediatrics that year.
03:45This, I think would be the spring
03:47of 85 in my backyard in Hamden.
03:50This is a young Luciano Pavarotti
03:51you see over here at the time.
03:53And this is and and again, Michelle.
03:56Jeff's daughter was not in this picture,
03:58but I've had the the pleasure
04:00of watching Doctor Gruen.
04:02Just rise through the ranks over the
04:04years and grow gradually to be the the
04:07fellow that you all know and love today.
04:10He became a professor of medicine.
04:12Excuse me of Pediatrics.
04:13In Unix in 2010 he's also an
04:16honorary professor at Shenzhen
04:18Tong University in China.
04:20He's internationally recognized as a
04:22scholar and investigator with collaborators
04:24around the collaborators around the world.
04:27He's very, very well published.
04:29He's been very well funded for
04:30a number of years, many,
04:31many years, and at Yale,
04:33he's really a valued mentor and advisor
04:36to medical students to graduate students,
04:38to, to our trainees, to our junior.
04:40Faculty and, and frankly,
04:42to at least one very grateful Chief for
04:45whom he's been a valued advisor for many,
04:48many years.
04:48So it's been my pleasure to watch Jeff
04:51really become such an incredibly,
04:53highly respected scholar and
04:55investigator that he is.
04:56You know,
04:57he studies genetic variants that
04:58underlie common learning disabilities,
05:00and you're going to hear a bit
05:01about that today.
05:02And as such,
05:03the talk should really be of interest
05:04to anybody who's interested in
05:06genetics or studies genetics or
05:07who wants to learn more about it,
05:09or anybody who's who works with
05:11kids with learning disabilities or
05:13wants to learn more about that.
05:14But there's the lecture will also
05:16be of interest to anybody who wants
05:18to be a better teacher,
05:19because as most of you know,
05:21and many of you know,
05:22and you're all about to find out,
05:24our friend Jeff can teach.
05:26He can take difficult subjects
05:27and make them accessible to
05:29to the point where even a bioethicist can
05:32understand he's an extremely gifted teacher,
05:34and he's also, by the way,
05:35a very gifted neonatologist.
05:37He is the triple threat that
05:39people talk about on academics,
05:41and it's really been my
05:42pleasure and honor to.
05:43Have this journey with Jeff for
05:45the last many years.
05:46As you can see from that picture,
05:47you know I think Cliff is still celebrating.
05:49We're all still celebrating 100 years
05:51of Yale Pediatrics and so and so they
05:54wanted a couple of guys today who
05:56actually were here 100 years ago at
05:58Yale Pediatrics and so here we are.
06:01It's an absolute pleasure
06:02to introduce my friend,
06:04Jeff Gruen.
06:14You know if you ever want an emotional
06:18experience, be introduced by somebody
06:20who's been really your friend,
06:22not just a colleague for 35 years, it's.
06:26You can really get choked up about
06:27this stuff. Thank you for doing that.
06:29Right before I have to give a talk,
06:30I really appreciate that Mark. Thank you.
06:34Glad to help, Jeff, thank you.
06:38So I'm going to talk a little
06:40bit about our new program for
06:43learning disabilities research.
06:44I'm going to basically answer the
06:47five basic questions, why, what,
06:49how, where and who very simply.
06:52Why is early identification of
06:54learning disabilities important?
06:56What is the scope of the problem?
06:57How will early identification disrupt and
07:01change the current educational paradigm?
07:03Where can we make a difference
07:05and who are we?
07:07So let's get started and with all our
07:10research we start in the classroom.
07:12And that's because worldwide,
07:14about 15% of students everywhere
07:16struggle with learning disabilities.
07:18These include things you've
07:20heard of and things you haven't.
07:22Most common is dyslexia,
07:24also known as reading disability,
07:26specific language impairment,
07:27which is delayed onset of speech,
07:30verbal trait disorder,
07:31which is impairment of articulation,
07:34dyscalculia,
07:34a math disability and attention disorder,
07:38is better known to most folks.
07:39Is ADHD and ADD dot.
07:42But what I'm talking about
07:43today and every and all the kids
07:45that I'll be talking studies,
07:46I'll be talking about today are students
07:49in regular classes at regular schools.
07:52And with normal or above normal intelligence.
07:56For most children, a durable
07:58intervention can make a huge difference,
08:01and I'm going to show you that
08:02data in a in a few minutes,
08:04but in the meantime there are three
08:06critical things to know about reading.
08:09The first is that dyslexia or
08:11reading disability is really common,
08:14so if I take the 2000 fourth
08:16graders here in New Haven,
08:18and I take them,
08:19I give them any standardized
08:20reading performance assessment.
08:22Let's say comprehension or single
08:24word reading, and we.
08:26We graph them the number of
08:28readers at each point,
08:29low scoring versus high scoring.
08:31Not surprisingly,
08:32you'll get a normal distribution
08:33if you have enough students with
08:35a mean and what we call reading
08:38disability is basically a cut off.
08:39It's a cut off at the tail end.
08:41It's a lower end,
08:42so most places do around the 10th percentile.
08:45Some places do 15, some do 12 percentile.
08:48But in the United States we're talking
08:50about 5 to 10 million US schoolchildren.
08:53Again, these are in regular kids.
08:54These are regular kids in regular schools.
08:58The second thing to know is
09:00that it's genetic, and so again,
09:01if we take our distribution of children,
09:04let's say again 4th graders and we do
09:06say comprehension and we chart them.
09:08So we have a normal distribution.
09:10Not only will it have a mean,
09:11but it'll be a variance around
09:13the mean as well.
09:14And if we ask the factors,
09:15ask which factors can account for that
09:18variance and in fact non genetic factors.
09:21To the surprise of most people actually
09:23account for a very small percent,
09:25probably in the 20 to 30% range.
09:27Ras cumulative genetic factors
09:29account for the vast majority of
09:32reading performance or the variance
09:34around the mean for all of us,
09:36and that's why it's really important
09:37to pick our parents carefully,
09:39because this is what really is
09:41important for our performance
09:42in reading.
09:45So what we say is that genetic factors
09:47account for up to 80% of reading performance,
09:50and we call this concept heritability.
09:52And since genetic factors are mostly
09:55responsible for dyslexia and our
09:57performance and meeting our past
10:00was to discover reading genes.
10:02And so to identify these kids at
10:04risk when intervention works best,
10:06we developed a genetic screening panel.
10:09We collect saliva from saliva.
10:11We extract genomic DNA and we've
10:13done these studies in two forms.
10:15Our first one was about 10 years ago.
10:17Was the grad study of the genes
10:19reading and dyslexia study.
10:20The first study of its kind of a
10:22strictly African American Hispanic
10:24American kids in the United States
10:26using genetics and then more recently
10:28the longitudinal program called
10:30the New Haven Lectionum project.
10:32And these have been informative for kids
10:35that are European American background,
10:37Hispanic American background
10:39and African American background.
10:41And we published these in in
10:43in in peer review journals.
10:45Some pretty high quality peer review
10:47journals and that in a very nice very
10:50nicely that many of them are reproduced
10:52sometimes in other populations,
10:54even in other languages.
10:57So the third thing to know
11:00about reading is that.
11:02Is that dyslexia can be effectively treated,
11:04and so let me share with you data
11:07from my colleague Maureen Lovett
11:09at the University of Toronto and
11:12State Kit and Maureen study.
11:15Does randomized control trials and
11:17kids who struggle with reading and she
11:19does a very intense intervention that
11:21lasts for about a year and what she's
11:24done is she's retested them two years
11:26after completion of the intervention.
11:28So these are outcomes 2 years after
11:30completion of an intervention for
11:32children that were picked up in grade.
11:34One children in grade two children
11:36in grade 3/4, etc.
11:38All the way out to grade 12 and what
11:41you can see here pretty obviously is
11:43that these are the kids that perform
11:45the best that are most responsive.
11:47So about 75% of these kids will
11:50respond to intervention and will be
11:52reading at grade level two years
11:54after completion of the intervention.
11:57Whereas in contrast,
11:57if you look at kids that are
11:59picked up in high school,
12:00which is about 50% or so of kids in
12:02the United States are picked up.
12:04Was reading disability in high school
12:07or later only about 1/4 of those
12:09kids will actually be respond will
12:11be reading it ever brought up to
12:13grade level or be reading at grade
12:16level even two years afterwards,
12:17so early identification is
12:19obviously really important.
12:20That's the key.
12:22The key to effective intervention
12:24is early identification,
12:26but the window of opportunity is narrow.
12:29So let me show you why.
12:31Here's our current approach in
12:32our educational system today,
12:34and that's in private schools,
12:35public schools everywhere we
12:36call it the wait to fail model.
12:38You'll be very familiar with this,
12:40and so this would be typical development of,
12:42say, comprehension.
12:43But again, it could be single word reading,
12:45could be spelling of kids of
12:47typical children,
12:48the United States and around the world.
12:50Frankly, in a pre K all the way up through
12:52grade five you can see it's nearly linear.
12:54The increase in the performance
12:56from year to year.
12:58But there are children who begin to
13:01fall off once they are come face to
13:03face with a challenging curriculum,
13:05so this will usually happen
13:06at the end of grade one,
13:07definitely in grade 2 where
13:09they really begin to fall off.
13:11In the current model,
13:12what we do is we wait until
13:14they're really performing poorly,
13:16and then we do a standard intervention.
13:18And even if the intervention
13:20is intense and even if they're
13:21young enough to respond to it,
13:23they never really catch up,
13:25and so this gap is never really bridged.
13:28What we're proposing is a different model.
13:30What we'd like to do is identify the
13:33kids here early and then track them,
13:35and then when they even fall off,
13:37or even even just mildly,
13:40then we can initiate an early intervention,
13:43and which they would catch up,
13:45or at least attain very nearly
13:46Gray level reading.
13:47And that is then sustainable.
13:49But the key here is the early identification.
13:54So in in order to do that, that's where.
13:57The the identification of genes,
13:59genetic variants and creation of a
14:02genetic screening panel is so important.
14:04So how do we do this?
14:05How do we do this magic of this
14:08early identification?
14:08And so as an example for a minute,
14:10let me just talk to you about
14:12some of the work we've been doing
14:14over the last seven years on
14:15the New Haven Leccinum project.
14:20You know, hey, the goal in New Haven
14:22election project has been to explore
14:24the extent to which genetic variants
14:26correspond with response to intervention
14:28through whole genome sequence analysis,
14:30serial behavioral testing and
14:32serial functional MRI studies.
14:35It's a longitudinal study of
14:37normally developing children.
14:38Grades we would enroll children
14:40in grades one and follow them
14:42all the way through grades five.
14:44We recruited these children from
14:4632 New Haven public schools.
14:47These are the regular schools,
14:49not the schools that have children with
14:52intellectually impaired children in them.
14:53We recruited about 500 and
14:55following about 500 children,
14:57these included a parent questionnaire
14:59based on for past medical history in depth
15:03questions about socioeconomic status,
15:04learning disability histories,
15:06primary language and spoken in the home, etc.
15:09And then for each child we did 30X
15:13whole genome sequencing through saliva.
15:18We also did standard psychometric
15:20batteries that assess reading
15:22language executive function.
15:24Beginning in grade one roughly every
15:26six months, all the way through Grade
15:285 and the people who did the executive.
15:30These assessments were folks from our group
15:32and people who are trained to do this,
15:34so we didn't rely on school testing.
15:36It was our own group who
15:38actually did the testing,
15:39and we used standard batteries.
15:41So just for giving an example for those
15:43of you who are familiar with the field,
15:45these are standard neuro
15:47psychometric testing batteries.
15:48These are not batteries that we
15:49made up or tests that we made up.
15:51These were chosen by skilled
15:53neuropsychologists to be part of
15:56the educational assessment tool.
15:58They include the tower two or the test,
16:00averting reading efficiency subtests
16:02from the Woodcock Johnson and subtests
16:05from the Grey Oil reading test,
16:06and so we actually combined
16:08for the first study.
16:09I'm going to show you to this today is.
16:14It's what we call a composite of five
16:16subtests from these standardized testing,
16:19the first one is a sight word
16:21efficiency from the tower,
16:22where we ask children in a timed manner
16:25to read real words like cat, dog, etc.
16:29And then we score them for accuracy.
16:31They do it for about 45 seconds,
16:33so the next one is a
16:35phonemic decoding efficiency.
16:36It's called and so and so that
16:38memory doesn't play into this.
16:39We actually give them non words.
16:41Words like zoup.
16:42Again, it's timed and we score for accuracy.
16:46The letter word ID subtest Woodcock
16:48Johnson also looks at real words.
16:51It's not timed,
16:52but it's progressively more difficult,
16:54and here we score for accuracy.
16:56The number of words they accurately
16:58get and then the word attack,
16:59which is again back to non words again to
17:01to really look at their ability to decode.
17:04Also non timed and finally to
17:06look at reading fluency you have
17:08them read age appropriate passage.
17:10It's time we score for accuracy
17:13but also comprehension.
17:14So we combine these five subtests.
17:16Into a composite phenotype.
17:19And it's quantitative, that is.
17:21We could score it when we score
17:24it by a number,
17:25so we have high performers
17:27and low performers.
17:28And then the the trick here is
17:31to relate it to the genetics and
17:33we can do that using within the
17:36same children single nucleotide
17:38polymorphisms and just to re familiarize
17:41yourselves with what exactly a
17:43single nucleotide polymorphism is.
17:45I put up this little diagram that is
17:49diagramming a specific single nucleotide.
17:52Polymorphism has this horrific name called
17:58Rs 6935076. It's specifically
18:00located on chromosome 8 and its
18:03position is 24 million, 644,000,
18:07zero and 94 nucleotides from the top
18:12of the short arm of chromosome 8.
18:15And here you can see the sequence
18:17is very specific and because
18:18all of us carry 2 chromosomes,
18:20there are two representations of this area
18:22and of this specific position and person.
18:25One here is inherited from his mother
18:27ASI and from his father he's inherited.
18:30Also, I see and so we would say
18:32that for this snip at this point,
18:34which is which has a location
18:37specifically assigned to it,
18:39this person is homozygous for
18:41the C allele or CC.
18:43This person I a different person actually
18:47carries AT inherited from the mother
18:49and AT inherited from the father.
18:51So therefore this person would be homozygous
18:54for the teal or TT at this specific snip,
18:57and then finally we have the heterozygous
19:00individual who inherited AC from one parent.
19:02And a T from the other parent.
19:03So this person would be called the CT or
19:06the genotype would be a CT designating
19:09that that this person has one of each.
19:12So that's what a snip is.
19:13It's just a single nucleotide difference
19:16at a very specific location for everybody.
19:19So Rs.
19:226935076 that specific location can
19:24be identified in all of us as well
19:27as our 500 children enrolled in
19:29New Haven election on project.
19:31So it does take a look at these
19:33three at these three genotypes.
19:36The CC,
19:37the see T and the TT and what we
19:39can do for any individual child
19:41since we now can determine the
19:43CCC T or TT status of that child
19:46is then related to performance on
19:49the decoding composite phenotype.
19:51And So what you can see here is
19:53that the CC in in in a single
19:56individual might my correspond with
19:58low performance and asked sort of
20:00in the middle performance and AT&amp;A.
20:03Why, in other words,
20:04it would appear here is that the
20:07C allele is detrimental and that
20:09the more see you have,
20:11the lower your performances
20:13in this decoding composite.
20:17So there's a relationship here.
20:18It's actually a linear relationship.
20:21More see the worse you do Morty,
20:24the better you do,
20:25and so if you do this over a lot of
20:28individuals then you can actually
20:30get a coefficient of correlation.
20:32The R-squared over those individuals
20:34and you could assign that value
20:37AP value a confidence value.
20:40And then you can graph that
20:42value for that individual snip.
20:44So here for this particular snip
20:486935076 located where I said 24
20:51million etc on chromosome 8 with a
20:54particular R-squared in 500 children,
20:56you can graph that on a P value
20:59on his P value.
21:00The confidence interval
21:01for that relationship.
21:02In fact,
21:03you can do this for a lot of snips
21:06on chromosome 8. Not just one.
21:10In fact.
21:11Every one of these dots represents
21:13a different snip genotyped in
21:15all 500 of our children.
21:17In fact,
21:18there are many snips on chromosome 8.
21:20There's well over 10,000, and so after.
21:23If you want to present them graphically
21:25again relative to their P value
21:27for their R-squared, you literally
21:29saturate or paint that chromosome.
21:32So represented here is every snip that we
21:35tested against the composite phenotype for
21:38decoding within our cohort of 500 children.
21:42And in fact we did this for every chromosome,
21:45so I just depicted chromosome 8.
21:47But this is chromosome 9,
21:49chromosome 1011, etc.
21:50All the way out to 22 and you'll
21:53notice that one is the widest because
21:56chromosome one is the longest,
21:58so it has the most number of snips
22:01and chromosome 22 is the shortest.
22:03So has the least number of Snips.
22:06This is real data.
22:07This is the real result we call this
22:09for obvious reasons in Manhattan plot.
22:12Right, and here's a peek at
22:13the Manhattan plot.
22:14In fact, for the decoding composite,
22:17we got a peak right here on chromosome 19
22:20by a number of snips in a very small area.
22:24And because of the magic of
22:25the Human Genome Project,
22:26we now can tell since we know precisely what
22:29the nucleotide number in chromosome it is,
22:32we can identify the gene
22:34that this corresponds to,
22:35and it corresponds to a gene that was
22:37not previously implicated in reading,
22:39but which we found to be very strongly
22:41have a very strong effect size called Gary.
22:45One in our cohort of 500 children.
22:48What do I mean by that?
22:49Let's go back into our cohort of 500
22:52kids and let's look at four different
22:54assessments of reading or reading subtests.
22:57The word, letter ID, word attack.
22:59Which word letter ID is timed,
23:02word comprehension, and reading fluency?
23:04That's the.
23:05That's the paragraph passage.
23:07And let's look at kids who have the
23:09Gary run risk allele here in orange,
23:11and you can see is that they enter
23:14first grade behind the kids in the
23:17same cohort at the same schools.
23:19At the same age in the same grade,
23:22at at, at, at, at,
23:24at a significant disadvantage.
23:25And if we follow these same kids
23:27because it's a longitudinal study,
23:29you can see they never catch up.
23:32Let me show you what that looks like.
23:34This is a comprehension,
23:36right,
23:36a fundamental and important things
23:38skill in reading.
23:39If we look at kids in grade at the
23:42start of grade one at the end of Grade
23:441 starter grade two end of grade 2 all
23:47the way up to the start of grade five,
23:49you can see is that children in oak.
23:51Our cohort here at the risk
23:53allele for Gary One,
23:54started off at a disadvantage
23:55to kids who do not carry any
23:57risk alleles for Gary one,
23:59and it's never really.
24:00They never really filled a gap.
24:02Even at the start of the Great Five,
24:04so it's sustained.
24:05It's a sustain effect over this age group.
24:12We can look further into Gary one we didn't
24:14know this so we looked in the text portal,
24:16which gives you expression data in different
24:19parts of the body and specifically
24:21in different parts of the brain.
24:23And to our surprise, Gary went well.
24:25It's not surprising that Gary
24:26wins expressed in the brain,
24:28but its strongest expression is
24:30in the cerebellum, the cerebellum
24:32hemisphere and total cerebellum.
24:35That was a surprise.
24:36But not really that much because what
24:38we know is that the cerebellum is
24:41intimately linked to the cerebral cortex,
24:43the cortical area where there are
24:47specific and specified specialized
24:49reading areas by a circuit or a
24:52cerebellar cerebral cortex loop.
24:55This is well known and in fact
24:57its support of a theory that was
25:00put forward in the 1990s called a
25:03cerebellar theory of reading disability.
25:05Nicholson Faucet observed that reading
25:07disability was also associated with
25:10deficits in cerebellar related functions.
25:12Distonia time estimation skill
25:15optimization and balance,
25:18and so they posited that the cortical
25:21cerebellar circuit shown here has
25:23an important role in reading.
25:24There has now been a whole lot
25:26more published.
25:27On the cerebellar theory of reading.
25:29And sometimes some people are in
25:31favor and some people are not,
25:32but we go where the science is taking us.
25:35And since Gary one is implicated,
25:36certainly by association studies,
25:39we pursue that.
25:41And so I don't really have the time
25:42to go through all the data and
25:44all the projects that we're going,
25:46so let me sort of summarize
25:49those briefly here.
25:50So our New Haven,
25:51Lexington project,
25:52which began to implicate the cerebellum,
25:54actually spurred a number of
25:57interesting projects Haley Dasilva
25:58in our lab has actually worked
26:01on a a program for processing and
26:04preprocessing imaging data that
26:06she calls Neuro Stack specifically
26:08made for a WS to Amazon Web server.
26:11And that's really important because
26:12a lot of the imaging data,
26:14for example from ABCD, is stored,
26:17allocated,
26:17processed and worked on in an
26:20Amazon Web Services up on the web.
26:22So in this is all cloud based work
26:24and Haley actually created this
26:27new tool that is now generally
26:29available as it's free and it's
26:32incredibly useful for neuroscientists.
26:33You know who don't want to get into
26:36the specifics of preprocessing and
26:38processing data. Tremendous work.
26:41We also calcium xerac in our lab who's
26:44been working with Young Frieder is unusual.
26:46Haley also,
26:47as well as Emily Curtain,
26:48has been working in ABCD.
26:50ABCD is a longitudinal study of
26:54typically developing 10,000 US
26:56children from across like 32 sites
26:58in the United States and they're
27:01getting cereal imaging studies
27:02as well as genotyping.
27:04Genotyping,
27:05one serial imaging studies and and
27:07as and some cognitive assessments
27:10to the NH Toolbox.
27:11That data is publicly available and
27:13so with kelson's been doing is he's
27:16been doing genetic correlation studies
27:18between the Leccinum project and ABCD,
27:20and in a way that he's been
27:22using these to create sort of to
27:24leverage our small data set with
27:27a larger data set. This is a theme
27:29that that goes on and on with us.
27:33Seeing Wang has been working by doing
27:36our due osis one I just showed you as
27:40well as T Ross is Tiwa Sizarr transcript
27:43transcript own wide association studies
27:45and So what she does is she goes back
27:49into GTX that expression data data
27:51data set in the brain either in the
27:53celebra ballimore the whole brain
27:55and she uses that information to
27:57wait the snips for the for the guasa.
28:00Therefore it's called that
28:02waser transcriptome.
28:03Wide Association study and she's and and
28:06we've been working with helping as well.
28:09Helping Zang Kate Connors from the
28:11group with young Frieders has been
28:13doing looking at our math data
28:15because I think initially,
28:17when I explained how the
28:19election project works,
28:20we do executive function which
28:22includes attention as well as math,
28:24math subjects as well.
28:26So Kate's been been really focusing
28:28on math and we'll be hearing some more
28:30data from that in the next couple of years.
28:33And then finally Steven Penny Agra.
28:35His spirit had had a collaboration
28:37with Indian Park over the stem Cell
28:40center and developing neurons human
28:42neurons from human embryonic stem
28:44cells and so the idea here being
28:46is that one of the things that's
28:47really held back.
28:48Translational Neuroscience is access
28:50to brain human brain material,
28:53and so unlike cancer,
28:54when you can get lots of material,
28:57in fact it's would be difficult.
29:00I would say it's impossible to get.
29:02Human brain material from from normal brains.
29:05And so in order to.
29:07If you want to study human neurons
29:09that what you do is you get human
29:12embryonic stem cells and then
29:14you begin to develop you.
29:18You can induce them to develop
29:20into human neurons along the way,
29:22and what Steven has done is
29:24he's knocked down using CRISPR.
29:26The expression of a prominent reading
29:31or dyslexia gene called KA 319 and
29:33what he's shown is that this puts
29:36the cells in a quiessence state,
29:38so these are neuronal progenitors
29:40arrested in a quiescent state.
29:42So to analyze that data which you know
29:45we've been doing a lot of sequencing there,
29:47is you.
29:48You are a few,
29:49I'm just doing a marvelous job
29:51again from Helping's group and one
29:52of our undergraduate students.
29:54Is he Lopez?
29:54So these are things that are
29:56currently ongoing.
29:57Major projects in the lab.
29:59I don't have time to talk about a
30:01whole lot today, but I just wanted
30:03to know that they're going on.
30:05OK,
30:05so the question is or what the
30:08question I posed before I left this
30:10slide was how do we do this magic?
30:13How do we identify these children
30:14early and when I said was we do this
30:17molecular genetic studies and I showed
30:19you a nice example focusing on dyslexia.
30:24But do we really disrupt by doing this?
30:27Do we really disrupt the current
30:29current educational program?
30:30Like how to disrupt the
30:32current educational paradigm?
30:33Is that really the question or is the
30:36question is how do we really disrupt
30:39the current educational paradigm?
30:40And I think that's really,
30:41really an important question for us,
30:43because doing our studies
30:45and publishing them,
30:46even publishing them in peer
30:48reviewed journals has been great.
30:50It's been great. You know,
30:51we've been able to get support for that,
30:53both private foundation.
30:54Age foundation and do
30:55these really cool studies,
30:57but are we really moving
30:59them into the classroom?
31:00And I would say I have to
31:02admit in fact that we were not.
31:03So if we really want to do this,
31:05where can we make these changes
31:07so we can translate our findings?
31:09Our ability to identify kids at
31:11high risk for learning disabilities
31:13at a young age when it's really
31:15important saying kindergarten?
31:16How do we get that adopted in
31:19public schools across America?
31:20Even in New Haven,
31:22across the state of Connecticut or America?
31:25So to do that,
31:27we really need policy level change.
31:30In the last five years,
31:31dyslexia and learning disability
31:33policies at the state and federal
31:36levels have improved significantly,
31:38most notably in the areas of
31:41teacher training and universal
31:42dyslexia screening in grades,
31:44kindergarten through third grade.
31:4841 states today have universal
31:50evidence based screening for dyslexia,
31:53including Connecticut and Massachusetts.
31:56And the feds have clarified that you
31:58can now use this term Once Upon you couldn't.
32:00You could use this term when,
32:03when,
32:04when teachers create the iep's,
32:06the individual educational
32:07programs that are so important in
32:10mandated for each child that has
32:12a disability learning disability,
32:1432 states across the country provide a
32:17list of approved evidence based screeners.
32:20None of them are genetic.
32:23So our goal is to pass legislation
32:26that includes reliable and validated
32:28genetic screening for risk of, say,
32:31dyslexia or any learning disability.
32:34And get it,
32:35get it on the improved list
32:37of evidence based screeners.
32:40This would allow parents to screen
32:42their child as early as even birth,
32:45but definitely pre K for risk of
32:46any of the learning disabilities,
32:48including dyslexia.
32:49It would require school districts
32:51to take those results into account
32:54when when they create the iips,
32:56the individual educational
32:57program plans and the 504 plans.
33:02And this would enable children
33:04finally to receive an evidence based
33:06reading instruction or intervention
33:08method at the very earliest
33:10stages of reading acquisition,
33:12when they are most important
33:14and are most likely to work.
33:17So I've last 20 minutes or so.
33:20I presented a hodgepodge really
33:22of information from a number
33:24of seemingly discordant fields.
33:26I presented some epidemiological data
33:29showing what is the frequency of of.
33:31Well, I presented reading disability,
33:33but in fact of learning disabilities and
33:35it's somewhere in the 10 to 20% range,
33:37I presented an intervention
33:39program from my colleague or or
33:41the results of an intervention
33:43intense intervention program for
33:45my colleague Maureen.
33:46Love it at the University of Toronto
33:48and showed how important it is.
33:50For early intervention and for
33:52it to work and to be and for
33:55its effects to be sustained.
33:57I presented some genetic genetic studies,
33:59both molecular genetic studies
34:00and statistical genetic studies
34:02to show the importance of
34:03genetics and why genetics should,
34:05could and should be used for early
34:07screening to identify kids at high
34:09risk for attention disorders,
34:11reading disabilities,
34:12language disorders, etc.
34:15And I presented an education paradigm
34:17that doesn't work the wait to fail role
34:20model and an alternative to that model.
34:22And finally I I presented some
34:25policy stuff and so the fact is
34:28is that it's a bit of a mess,
34:30so to find a form that accommodates
34:33the entire mess that I've just
34:35presented and to unite these from my
34:38friend Samuel Becker from 1961 and
34:40to unite these seemingly unrelated
34:43initiatives under a single umbrella
34:46this year.
34:48With a lot of help from Gene Shapiro,
34:50Cliff Boag and the Dean,
34:53we created the Yale Program for
34:57Learning Disabilities Research.
34:59Our program is multidisciplinary.
35:01It covers the fields of neuropsychology,
35:04Biostatistics,
35:05education, neuroimaging,
35:06and genetic and the people
35:09who come into this group.
35:12They know their field well,
35:14but neuro psychologists who come in
35:15don't know a whole lot about genetics.
35:18Geneticists that come in don't
35:19know a whole lot about education,
35:22often not a whole lot about neuroimaging,
35:24so it's some very important and
35:26we really stress that we begin to
35:28learn each other's vocabulary that
35:29we really interact personally.
35:31Although this year it's been
35:33difficult to interact personally and
35:35so we can inform each other studies
35:37and help and work collaboratively
35:38collaboratively within the lab.
35:40And I think we've been pretty successful.
35:42I think that's like,
35:43for example,
35:44the study I presented from the
35:45Haven election on project.
35:46You can see that's a combined effort from
35:50neuropsychologist geneticist Biostatistics.
35:52I didn't show you the neuroimaging data,
35:53but it's remarkable how well that works and
35:56now we're moving into education with a PhD,
35:58PhD,
35:59postdoc in education recently joined the lab.
36:01So these are great initiatives
36:03and show how we're really blurring
36:05all the lines amongst all these
36:07different disciplines.
36:08But we're also trying to go cross
36:10campus and reach out and bridge
36:12the gap to other schools to other
36:14students and the other professional
36:16schools in the university.
36:18So I'm especially proud to say that
36:20we've been working with the folks at
36:22the Yale Educational Studies program,
36:24and we'll be offering a A1
36:27semester long course,
36:28mostly for undergraduates
36:29beginning this fall.
36:31This is the first course that they've been.
36:33Offering on learning disabilities and
36:35it's a full I think it's 17 lectures
36:37over the course of the semester,
36:39so this is developed by the
36:41folks in our in our group most.
36:44There's no Lee,
36:46Kim Tsujimoto and John Boston,
36:49and so they're prepared to present this.
36:51It is currently under review, but it's it's.
36:53It looks like it's going to start
36:55in the fall. We work really strongly
36:57with the folks in the computational
36:59biology and bio for Mattix program,
37:01but they have members from those
37:03programs in our lab and plus we
37:05present to them on a regular basis.
37:07This is a really extraordinary new
37:09program here at the School of Management.
37:12It's called the broad center or
37:14the Broad Center for Educational
37:16Leadership and Public Education.
37:18It's a brand new initiative.
37:19I think there are about a dozen broad
37:21Centers for this across the country.
37:23It was just started this past
37:24fall that brings in educational
37:26leaders from around the country.
37:28These are superintendents and school
37:30principals in the largest districts in the
37:32United States and they engage in a yearlong,
37:34transformative leadership
37:35program in public education.
37:38And so we're in contact and we're
37:39working with these folks as well.
37:41And then finally,
37:42of course,
37:42we're based here in the Department
37:44of Pediatrics and Genetics,
37:45where which is our home base and
37:48where we're going to also be reaching
37:51out to residents and hopefully.
37:53Increase the exposure they're
37:54having to learning disabilities
37:56that they're going to encounter
37:58when they enter practice as well,
38:00'cause it's so common.
38:02So this is our Yale program for
38:05Learning Disabilities Research.
38:06Its comprehensive, it's large,
38:08it's ambitious.
38:10We just got started this past September.
38:12These are the different components
38:14I tried to present to you,
38:15and with that I'm happy to take
38:18questions and also a big shout out
38:20to all these wonderful people who
38:21make this sort of research possible,
38:24as well as our funding groups
38:26from private foundations,
38:27them and foundations to the
38:28NIH as well as to individuals.
38:31And of course, thanks to all the children.
38:33Families who participated in
38:34all these studies.
38:36So if there are questions,
38:37I'm happy to take them.
38:44OK, Jeff, thank you so much.
38:46That was amazing.
38:47That was fantastic.
38:48Now we do have questions.
38:49I mean, I invite folks to submit
38:52questions via the chat or the Q&amp;A,
38:54but we've got some questions already.
38:56Doctor Who in one of your
38:58earlier slides indicated that
39:00heritability for dyslexia is 0.80.
39:02How much of that is accounted
39:04for it by the SNP and Gary one?
39:06Is there a polygenic risk score for dyslexia?
39:09If so, how much variance?
39:11Does it account for?
39:12That's a great question.
39:14Thank you for asking so as clearly
39:17the you know Paul knows is that
39:20there's really two types of.
39:23Yeah, of variants that you can account
39:25for heritability that you can account for.
39:27There's broad sense heritability,
39:29and there's narrow sense heritability.
39:31When I say 80%, it's it's broad sense
39:34heritability that is when you look
39:36at twin studies and family studies,
39:38and then you see,
39:39and then you ask the question,
39:40what's the concordance rate between
39:42siblings who are affected, say,
39:44was reading disability in those that aren't,
39:46and you compare them the the classic
39:48example would be identical twins
39:50which share 100% of their genome.
39:53The concordance rate there versus
39:55the concordance rates and non
39:57identical twins which only share
39:58on average about 50% and so if
40:00you do those sorts of those are.
40:02Those are standard heritability studies
40:04in the 1980s and they would say that
40:07the genetic component is about 80%.
40:10There has been a polygenic risk score.
40:14Publication from Robert Plomin's
40:17group from the UK.
40:19He has the largest collection of twins,
40:22I think in probably the largest one is
40:24in Australia but he has the largest
40:26one that was reading assessments.
40:28I think it's 10,000 twins and he
40:30was able to account for all snips
40:32that he looked at and he looked
40:34at a genome wide panel.
40:36I think about 700,000 snips.
40:38He was able to account for
40:40about 8% of the heritability.
40:42This is pretty common
40:44for polygenic disorders.
40:45That is that even if you look at cancer etc
40:48etc etc and just look at common variants,
40:51which is what's represented and a
40:53single nucleotide polymorphism panel,
40:55you generally there's a huge gap
40:57between broad sense heritability
40:58and narrow sense heritability.
41:00Now, to answer you Paul direct
41:02your question directly, Paul,
41:03if you look at any single snip,
41:05you're looking only at a fraction
41:07of the heritability that is Gary.
41:09One explains only a tiny amount,
41:12so all that's why.
41:13And I didn't get into it on this talk,
41:16but in previous.
41:17Oxide, I mentioned it is that is
41:19that dyslexia language acquisition,
41:21language impairment,
41:23verbal trait disorder,
41:25dyscalculia,
41:25or the all these are polygenic disorders.
41:28You have to be lucky enough to
41:30have the smorgasbord of the right
41:32variants all at once for that.
41:34For that to happen,
41:35it is rarely a single gene disorder,
41:38although we have been looking
41:40at rare variants,
41:41and it does occur from time to time.
41:43None of them are in Gary want.
41:47Thank you Jeff. A much more general question.
41:48How did you become involved in this research?
41:52Uh. So. Most people think that you know,
41:58Jeff. He's a pediatrician,
41:59and so he cares about children
42:01and he cares about reading and he
42:03got somehow seduced into genetics.
42:05But actually it's the opposite.
42:08And yes, I'm a pediatrician.
42:10Yes, I care about children.
42:11Yes, I care about reading,
42:12but the way I got in was during
42:15fellowship with with you Mark,
42:17I actually got first.
42:19Initially I was in cell biology and then
42:21I got seduced over to molecular genetics
42:23just at the time of the beginnings.
42:25Of the Human Genome Project,
42:27and I thought, you know,
42:27this is a pretty cool thing,
42:29and so I ended up in a lab that
42:31does that was really pioneering
42:33many of the methods for the human
42:35Genome project and my first.
42:37This was insuring Weismans lab here at Yale,
42:39and at that time what we were doing is
42:42remember there is no human genome project.
42:44It was just starting off the ground.
42:45Nobody really knew how to
42:47clone the human genome,
42:48and so one of my early
42:49projects was the clone.
42:50The short arm of chromosome 6 and then one
42:52day Sherm came up to me and said, Jeff,
42:55why do you clone the hemochromatosis?
42:58And of course, I said sure you know,
43:00I'm a pediatrician,
43:01I don't really know very
43:02much about hemochromatosis.
43:04It sounds like a disease of adults and
43:07Alcoholics, and he goes no, no, no you.
43:08We we've known that it's genetic
43:10for a long time,
43:11and you've already cloned it.
43:13It's one of your one of your 6 or
43:157000 tubes you have in your freezer,
43:17so just go figure out which one it was,
43:19and so that's how it got started.
43:21We got scooped by a private
43:23company several years later,
43:24but I had all these resources
43:26for chromosome 6.
43:27I looked around and there was this locus.
43:29There was this location of a strong
43:32genetic effect on the short arm
43:34of chromosome 6 for which I was,
43:36I think,
43:36at that time probably the only
43:38person in the universe who actually
43:40had complete coverage in new
43:42were all the markers were,
43:43and that's how I got started.
43:45And so the truth is, yes, I'm Peter, Trish.
43:48And yes, I practice medicine.
43:49Yes, I care about children.
43:51I care about reading,
43:52but the reason I really jumped into
43:54this is 'cause I had a strategic,
43:56scientific advantage.
43:57And I took advantage of it.
44:01Thank you. From Steve uptegrove.
44:05He even has, unfortunately a well
44:07known history for a high prevalence
44:08of childhood lead poisoning while
44:10childhood lead exposure is supposed
44:12to be documented on all children PE
44:14forms upon school entry in Connecticut.
44:16This information is not routinely been used
44:18to identify those at high risk for the
44:21same disabilities you were concerned with.
44:23Rather, they come to the light
44:25only by the same weight to fail,
44:27model the wait to fail model you mentioned.
44:29What has been lacking are the
44:31resources in schools for particularly
44:33districts like New Haven, Kubara,
44:35Heavy burden of potentially at risk hits.
44:39Steven, that's a great question.
44:41First of all, I want to tell and just,
44:43you know, reach out to Steve is that
44:45when we first started our studies
44:47here in New Haven Public Schools,
44:49Steven was one of the earliest
44:51persons that I reached out to because
44:54I think at that time Steve you
44:56were either on New Haven School,
44:58you were on the school board or you
45:00add connections to the school board.
45:02And so you were one of the first people
45:04and you were wonderful and I greatly
45:06appreciate all the help that you
45:08offered and all the advice that would.
45:09Proved to be very, very useful.
45:12Second of all is that I'm not
45:15discounting the environmental exposures,
45:17they're huge and and they're a big deal,
45:19and they are certainly a big deal
45:22to different socioeconomic groups.
45:23I can't answer your your your excellent
45:26question directly because that would
45:28that would require is probably you know
45:30more lead testing and more documentation,
45:33but what I can say is that we've
45:35begun as part of the Yale Program
45:37for Learning Disabilities Research.
45:39We've begun a formal program to look
45:42at the electronic medical record we
45:44started looking at by just looking
45:47at coding for learning disabilities.
45:49And we we're doing this project
45:52collaboratively with Emily Power,
45:53so so I think we're just starting to
45:55get the data at interesting enough.
45:58Interestingly enough,
45:58we're probably looking right now
46:00at about 40,000 kids in total,
46:03and amongst the 40,000 kids there.
46:05We're looking for codes that would at
46:07least implicate some learning disability.
46:09From there, we'll go on and that is.
46:11We'll try and peel the onion and
46:13get to be well.
46:14What is the real learning
46:15disabilities that they have could be?
46:16They could? Could they be overcoated?
46:18Could be undercoated,
46:19but one of the things that we
46:21will definitely look look for now
46:23that you've mentioned it.
46:24And thank you for asking is we'll
46:26look at lead levels 'cause we can do
46:28that in the electronic medical record
46:30and we'll try to see if that is a
46:31significant factor in these things.
46:34I suspect it is,
46:35but I also suspect there correlate
46:37strongly with socioeconomic status,
46:39which is something I didn't mention
46:41in the GWAS that I presented.
46:42But in that GWAS it is corrected
46:45for socioeconomic status.
46:46We always do that as well as sex and age,
46:50and so that's an important part.
46:51But I'm going to add lead to the list.
46:53I think that's a great question.
46:57Jeff from Jim Pelligrini Jeff.
46:58Truly amazing.
46:59You mentioned executive functioning.
47:01Any research at the genetic level.
47:02Here schizophrenics can have
47:04excellent verbal and reading skills,
47:06but severely poor executive function.
47:11No, I I've been very careful to
47:13stay away from the psychosis Jim.
47:16That's a that's a well populated
47:19field to be politically correct.
47:22And so I, you know, sort of stay in my
47:25lane and within within with children
47:27and typically and specifically.
47:29And typically developing children
47:31if you're not typically developing.
47:34That is, if you have a neuro
47:36psychosis or even autism.
47:38We've been excluding those
47:39children from our studies.
47:41Really trying to focus on those
47:43pathways and circuits that are
47:45specific for either reading
47:47language or executive function.
47:49Having said that,
47:51we have extensive executive function
47:53assessments on all 500 of our kids
47:56in our longitudinal study as well as
47:59the 1300 kids in our initial cross
48:01sectional study called the Grad study,
48:03and so we've actually performed
48:05our first juos and tiwas on that
48:07data specifically on attention,
48:09and I can tell you that our
48:11preliminary results.
48:12Look really,
48:12really good,
48:13so I think yes we will have some
48:16jeans and genetic variants that
48:18will correspond beautifully and and
48:20will identify risk for attention.
48:22I can tell you is that that there
48:24are a lot of good people in this
48:26field and a number of handful
48:27of genes for us specifically for
48:29attention have been identified,
48:31and so I think this is a very
48:33cool area of interest.
48:35We continue to pursue it
48:37and I'll keep you posted I.
48:39I think we'll have something solid.
48:42I'm hoping the manuscript will
48:43go out before June.
48:47Thanks Jeff from Julia Rosenberg.
48:49Thank you for a great talk.
48:51I'd love to learn if English language
48:53learners have been included in these
48:54studies in New Haven schools and if
48:56you have any insight into how some of
48:58these findings may factor in for those
49:00who are English language learners.
49:02Thank you. Hi
49:03Julia, so the short answer is yes,
49:06you know there are.
49:07So if you're going to be in New Haven
49:09public schools, you're going to be.
49:11You're going to have English
49:13language learners. So the.
49:15We for our initial study of the grad study,
49:19we required that so that
49:21was an older age group,
49:23so those kids were nine,
49:26roughly 9 years old,
49:27and we required that they have been at
49:30least three years in instruction in the
49:32United States to be in our studies.
49:34Remember that study?
49:35The grad study was the 1st,
49:37and remains the only study of
49:40underrepresented minorities of
49:41genetics and reading disability effort,
49:44and so, and we identified variants.
49:46Do that and publish variance through
49:48that and so that was for the grad
49:50studies in the haven next rental project.
49:53We didn't have much.
49:55Basically they could get by
49:57with a cursory understanding.
49:59In English we didn't.
50:02We didn't offer in the initial
50:04study the grad study we offered
50:06Spanish language testing.
50:08We actually, interestingly enough,
50:09we really see a difference between
50:12English and testing them in English
50:14or Spanish and the New Haven next.
50:16Project just about all the kids
50:18are pretty proficient in English,
50:20but it wasn't an exclusionary criteria.
50:24Thanks Jeff. Question from Susan Blvd.
50:28Do you know why New York and
50:30California were two of the states that
50:32were not doing dyslexia screening
50:33and any concerns that prenatal
50:35screening might be used to terminate.
50:37Quote non ideal End Quote babies so
50:41I can't tell you about California New York.
50:43Sorry but I do know that there that
50:46they haven't passed legislation
50:47yet for universal screening.
50:49I suspect it has a lot to do with
50:51money and so you know these are
50:53expensive things, but I don't.
50:55I don't know the specifics.
50:56Regarding prenatal diagnosis,
50:57I get this question a lot.
50:59People find me, they call me and
51:01this is what I tell them and it,
51:03you know I I don't work
51:06with children with dyslexia,
51:08I'm a neonatologist,
51:08but I my group does and and and
51:11certainly I interact with people
51:13from around the world that work with
51:16children with reading disability
51:18and they tell me two things.
51:19Yes, they struggle.
51:21And they also tell me that a
51:23significant number of them have
51:25other really significant talents.
51:26For example,
51:27there's a group that has told
51:28me that about 20% this is,
51:30and I've I've heard this from multiple
51:32neuropsychologists that there's
51:33probably somewhere between 10 and 20%
51:35of kids with really severe dyslexia.
51:37Have this special ability to be able to
51:40see things in three dimensional space.
51:43It's really a marvel of ability to do,
51:46and so if you if you look at all these
51:48examples of successes and people
51:49who have had pretty severe dyslexia.
51:52Done well business and academia
51:53etc and they they have been tested
51:56and they really do have dyslexia.
51:58The question is why would you
52:00want to exclude them?
52:01Why would you want to
52:02do an early termination?
52:03So I do my very best to when people
52:06call and I think I've been pretty
52:09convincing is that this would
52:10not be something that he won.
52:12I would either test 4 or #2 recommend
52:14that it be test four or certainly
52:17predicated termination on because
52:18it makes no sense whatsoever.
52:20These kids are.
52:22Are phenomenal,
52:23and they turn into phenomenal adults
52:25and incredibly happy and productive.
52:27They work hard, they struggle.
52:29There's no question about it and
52:31their parents struggle as well,
52:32but they do universally, really well.
52:36Well, I think that these kids also may
52:39have an even brighter future based on
52:40some of the work you've been doing.
52:42Jeff, I mean the the ability to
52:45identify these kids early and
52:46really change a lot of lives.
52:49I have a from auto Phoenix.
52:51Ask says this is fascinating and so
52:53so important that I'm thrilled that
52:54you were driving to policy with this
52:57for school aged children earlier than
52:59reading failure is language delay.
53:01Does any of the data point to a way to
53:03screen genetically for this in order
53:06to intervene even before school age?
53:08If we could do newborn screening
53:09for language delay,
53:10we could start early
53:12intervention immediately.
53:13So I should pay you out 'cause that was like,
53:16that's like the question about language
53:18delay or specific language impairment and
53:20reading disability comes up all the time.
53:22Roughly, if you if you if you if
53:26you do a really careful history on
53:28children with reading disability,
53:30about a third of those kids had had
53:34very significant language delay,
53:36that is delay in onset of language again
53:38just to remind everybody I'm talking
53:40about typically developing children.
53:42Children with normal IQ's, but if they're
53:44struggling and they clearly have,
53:46you know, and they're tested,
53:47and they really have a reading
53:48disability if you do a careful history,
53:50you'll find that roughly a third
53:52of those kids.
53:53Well, it had language impairment,
53:54delayed onset and speech.
53:5618 months, 22 months,
53:5724 months even older than that,
53:59there is a significant overlap and when
54:01we look at the genetic association of
54:04jeans that were primarily identified
54:06for reading and then you go around and
54:08you and you use those as candidate
54:10genes for looking at language,
54:11they overlap a great deal.
54:13There are some unique genes specifically
54:15for language impairment and there
54:16are unique genes specifically
54:18for reading disability,
54:19but the overlap is significant
54:20and that can explain why there's
54:23the shared heritability.
54:24If you go forward and you look
54:26at kids who are diagnosed with
54:28specific language impairment,
54:29again,
54:30about 20 to 30% of those kids will
54:32get into trouble by the end of first
54:34grade beginning of 2nd grade and will
54:36have reading disability so those kids
54:38have to be very carefully watched as
54:40far as newborn screening is concerned.
54:43When we ask the question a few years ago,
54:46how many tests in a newborn
54:47screening were done by DNA?
54:49The answer was zero.
54:51That's slowly changing,
54:53and so DNA screening.
54:55Not only is it potentially
54:57incredibly sensitive and specific,
55:00it's scaleable.
55:01That means the cost can be incredibly
55:04low and so that's what the hope is,
55:07and so we're pursuing various ways
55:11to get states to think about this,
55:13or at least to fund this at
55:15least as a pilot project.
55:16But that's exactly right auto.
55:18That's what we're trying to do.
55:21I think we have time for one more question.
55:23Jeff from Joe Abney.
55:24Singer exciting work.
55:25Have you moved the genetic testing
55:27to the point of the test that has
55:29a receiver operating curve that
55:30provides data on sensitivity,
55:32specificity, positive predictive
55:33value and negative predictive value.
55:36That's well, let me finish the question here.
55:39Hold on to that since kids with
55:40dyslexia are a heterogeneous group,
55:43does your work focus on the most common
55:45phenotype of phonological awareness,
55:46or do you have data on the
55:49orthographic variants?
55:51OK, so let me attack the first
55:52part of the question. First.
55:54Joe, it's a great question.
55:55Short answer is no,
55:57and and the reason is because
55:59in or certainly we could do it
56:01retrospectively and when we do
56:03we get a reasonable sensitivity
56:05and incredibly good specificity.
56:06But as far as a receiver operating
56:08curve you really want to do that.
56:10We could do that, you know,
56:13looking backwards,
56:13but the real way to do this would
56:16be a uh as a looking forward that
56:18is your identified or you just.
56:21You take a cohort of children,
56:22follow them for three to five
56:24years and and and do testing,
56:25and that's what the next stage is so we
56:27can do that receiver operating curve.
56:29So we're not there yet,
56:30but I think we're getting
56:32closer to being there.
56:33As far as subtests are concerned,
56:35this is the Holy Grail of neuropsychologists.
56:37I get asked this question all the time.
56:39Can you tell me?
56:40Does the reading disability
56:41that we're seeing in this child?
56:43Is it more to do with orthographic coding?
56:45Is more to do with phonological
56:47awareness and and and so then the
56:49intervention might be a little different
56:51and I can tailor to the child.
56:52We'd love to do that.
56:54The thing to remember is that
56:56there is no orthographic coding,
56:58and for the the the for the non
57:00informed of the group when we're
57:02talking about is really spelling
57:04there is no spelling gene.
57:05There's no grammar gene,
57:07there's no decoding gene.
57:09Remember genes encode proteins,
57:10proteins make up the receptors and
57:13the neurotransmitters or the brain.
57:15They make up the glea, the neurons, etc.
57:17The workhorses of the cell,
57:19and so they worked together
57:21in combination with different
57:22systems in order to enable us.
57:24To do these very complex very human
57:27things called decoding or reading
57:29and spelling and music as well.
57:32And so it it's not.
57:33It's it probably isn't going
57:35to be a single gene,
57:36but it may be a panel of genes
57:39or a panel a panoply perhaps,
57:42or a profile of certain variants
57:44that may be more weighted towards,
57:46say, decoding versus spelling,
57:48but we don't know yet.
57:50The studies aren't large enough yet.
57:53Well Jeff, you know you.
57:54You left plenty of time for
57:56Q&amp;A and our time is up,
57:58but I apologize to the many people.
58:00There's a lot of enthusiasm for
58:02this talk and many questions and
58:03comments which we didn't get to.
58:05So let me just congratulate
58:06you on a marvelous talk and I
58:09think that takes up our hour
58:10and thanks everybody for coming.