37 – Interpretation of SYNGAP1 Variants

Event Time

August 19, 2021

Dr. Perez-Palma is a bioengineer and a Doctor in Molecular Biosciences. After finishing his studies in Chile, he did a post-doctorate at the genomics center at the University of Cologne, Germany funded by the Dravet Foundation. Since then, Dr. Perez-Palma has specialized in the genetic study of patients with epilepsy and neuro-developmental disorders, focusing on the development of tools for the interpretation of variants. He was a research associate at the Cleveland Clinic until last year, and now he is at the Center for Genetics and Genomics, at the University of Development in Chile.

Here are our introductory comments: 

Hello everyone, and welcome to today’s session. My name is Marta Dahiya; I’m a Syngap parent and a Director of the Syngap Research Fund.

We are very excited to continue the SRF Syngap research fund webinar series. The goal of the series is to empower your communications with clinicians as you get more clear knowledge of SYNGAP. 

We also want to give a plug for our next presentation, “Treatments in development for epilepsy syndromes: opportunities for Syngap1”, which will take place on Thurs. Sept 2, 10 am PST/1 pm EST with Dr. Ana Mingorance.

Our talk for today is “Interpretation of SYNGAP1 Variants”. I have the pleasure to introduce today’s speaker, Dr. Eduardo Pérez Palma.

Dr. Perez-Palma is a bioengineer and a Doctor in Molecular Biosciences. After finishing his studies in Chile, he did a post-doctorate at the genomics center at the University of Cologne, Germany funded by the Dravet Foundation. Since then, Dr. Perez-Palma has specialized in the genetic study of patients with epilepsy and neurodevelopmental disorders, focusing on the development of tools for the interpretation of variants. He was a research associate at the Cleveland Clinic until last year, and now he is at the Center for Genetics and Genomics, at the University of Development in Chile. 

Webinar Overview

Dr. Eduardo Pérez Palma is a bioengineer and doctor in molecular biosciences at the Universidad del Desarrollo in Chile. He starts the webinar with the basic concepts of the human genome and the main types of genetic variants. He goes on to talk about how a third of SYNGAP1 variants are variants of unknown significance (VUS) which is a problem when trying to treat disorders. In order to interpret a variant and its relationship with a disease, there are established criteria for determining whether variants are pathogenic, benign, likely pathogenic, likely benign, or VUS. Dr. Pérez Palma compares this interpretation process to solving a crime. He then talks about the bioinformatic methods used to interpret variants and how variants are re-interpreted as new information is discovered. He concludes with how this field is beginning to make predictions of the functional consequences and the phenotypes beyond the pathogenic variant, providing more information to understand neurodevelopmental disorders.

Other Relevant Publications by Dr. Pérez Palma

Epilepsy Genetics and Precision Medicine in Adults: A New Landscape for Developmental and Epileptic Encephalopathies

Clinical sequencing yield in epilepsy, autism spectrum disorder, and intellectual disability: A systematic review and meta-analysis

THIS IS FOR TRANSCRIPT ONLY:

:03hello everyone and welcome to today’s session my  name is Marta I am a Syngap parent and director of  

0:10Syngap Research Fund we are very excited  to continue the SRF Syngap research fund  

0:17webinar series the goal of the series is to  empower your communications with clinicians  

0:22as you get more clear knowledge of syngap we also  want to give you a plug to our next presentation  

0:30treatments in development for epilepsy syndromes  opportunities for SYNGAP1 which will take place on  

0:37thursday september 2nd 10 a.m pacific time and  1 p.m eastern time with dr ana minguranzi our  

0:46talk today is interpretation of SYNGAP1 variant  i have the pleasure to introduce today’s speaker  

0:53dr eduardo perez palma dr perez palma is a  bioengineer and a doctor in molecular bioscience  

1:02after finishing his studies in Chile he did  a postdoctorate at the genomic center at the  

1:08university of Cologne Germany funded by the  Dravet foundation since then Dr Perez palmer  

1:15has specialized in the genetic study of patients  with epilepsy and neurodevelopmental disorders  

1:22focusing on the development of  tools for interpretation of variants   he was a research associate at the cleveland  clinic until last year and now he is at the  

1:34center for genetics and genomics at the university  of development in Chile a recorded version of  

1:42this webinar will be available on the SRF  website under webinars on the family menu  

1:49by the end of the presentation you will have  the opportunity to get the answers to your  

1:54questions we love to hear from you please write  your questions in the Q&A for those of you just  

2:02joining us welcome and again the speaker is  dr eduardo perez palma and the talk today is  

2:08interpretation of sync up one variant it’s now my  pleasure to turn things over to dr palma welcomePresentation Contents

2:18thank you marta it’s my pleasure to be here today  and thank you also for that kind introduction

2:26um well as previously announced my  the title of my presentation is called  

2:34interpretation of single single thing at one  variance and well the contents of this talk  

2:42first i’m going to introduce to you two basic  concepts about genetic variation and SYNGAP1  

2:50then i will continue with varying interpretation  concepts the methodology variant interpretation  

2:56and then and to describe to you some complementary  methods that we can input to this process and  

3:02help to aid in the identification of pathogenic  variation then i will continue with an example  

3:09of how to interpret SYNGAP1 variants  and then some ideas and concepts that  

3:16will try to describe to you where’s this  whole field going and future projections  

3:24so first let’s start with genetic  variation and think of one  Genetic variation

3:29genetic variation in the human genome is highly  normal between two individuals you can expect  

3:35between four and five five million nucleotides  involved in genetic variation among these any  

3:42normal individual with no apparent disease will  have between 150 and 200 protein 2 truncating  

3:48variants which these ones are usually diagnosed  as pathogen variants but naturally we have between  

3:54150 and 200 of these events also the most common  type of variation is by far missus variation  

4:05here we can expect normal individuals between ten  thousand and twelve thousand mistransparent events  

4:12of all of these genetic variants that  one will carry between 40 and 80 events   will be the novel that that means that it  will be absent from both of our parents  

4:23and among these one or two will be  on the calling portion of the genome   that means the genetic code that leads to protein  production these numbers tell us the first message  

4:36that any genetic test is going to find variance  this is going to happen the key question here  

4:43is which ones are the ones causing the disease  which are the ones that are pathogenic variation  Types of genetic variation

4:50we have many types of genetic variation we  can expect in every gene in our genome to have  

4:57at least two copies one from uh our mother and  another one from our parent and nissan’s parents  

5:04change a single nucleotide that leads to a change  in an amino acid within the protein sequence this  

5:13this may or may not leads to a chain of function  we have also protein truncating variants that  

5:21stop the protein production and then meeting  to usually leading to a an apple insufficiency  

5:28that means because we have only just one copy of  this gene also we can observe within the human  

5:34genome whole gene the lesions holding gene  duplications the functional effects of this  

5:41variation is very valuable too we have gain  of function and lots of function from missions  

5:47variance we have usually mostly loss of  function variation within protein through  

5:52the variance and from deletions also from  duplications we can expect gain of function  

5:58because we have multiple copies of the same gene  of course in biology there is an exception for  

6:04everything so if you go through the literature  you can find exception for each of these cases

6:12missions variation by far are the most  common type of variation and also the   most difficult to predict if we go a bit more  in deep into genetic variation and this is  Relationship between genetic variants

6:23we can plot uh the relationship between  genetic variants and this is in in this way  

6:29in the y-axis we have penetrance which is  how severe is the effect of the variant  

6:35and then on the x-axis we have the frequency  how common are variants within our population  

6:41so whenever a variance is is associated to  this use we can have a pathogenic effect  

6:48that is mainly populated by rare variation that  have a large effect these are the variants that  

6:54by themselves can cause disease these are very  rare they are not established in the general  

7:00population but they have a high effect then on  the other side we have common variations that  

7:08contribute a small effect to the to the disease  any disease and thus they can be establishing and  

7:14be common population these variants are thought  to have a collective effect a cumulative effect  

7:21so on this side on the genetic spectrum we can  start to talk about risk and an ethnic company  

7:29naturally uh developmental epileptic  encephalopathies and single one variation  

7:34falls within the realm of mendelian variation of  rare variants with large effects since one as you  Risk

7:43all know it has a complex presentation but it’s it  has a monogenic mechanism of inheritance usually  

7:51when the novel variant can cause the disease by  itself it has an incidence an incidence of one  

7:57every 16 000 uh naturally born individuals and  we can expect uh 400 new cases per year in latin  

8:07America and in the united states this number  it’s a bit lower is i think 366 individuals  

8:15new cases per year the presentation of the  disease is really complex we can have a lot  

8:22of co-morbidities and and symptoms within  SYNGAP1-related disorders and these symptoms  

8:28are complex in a time dimension too they they  evolve differently and they are they present  

8:34themselves in different ways as long as the time  moves and this is really important for the study  

8:41of these disorders to know to manage expectations  and and and see what we can expect from different  

8:48mutations it’s always really heterogeneous for  this type of syndrome but one important thing to  

8:56know from the beginning is that not all sync  up one variants will cause the same disorder  

9:03but also not all think of one variance will cause  a disorder by themselves most of the genes that  

9:10have been associated with neurodevelopmental  disorders they have naturally benign variation  

9:17within them and it’s easy to find variants that  do not cause any any harm at all and now in this  

9:27parent interpretation is key so let’s  move to the process of iron interpretation  Interpretation

9:34this is a typical genetic test and from a gene  panel you can see that we have 151 genes that  

9:40we know that are commonly associated to  disease and then the company tell us a  

9:46clinical interpretation in this case will have  a single one variant of uncertain significance

9:55this type of interpretation does not allow a  diagnosis it doesn’t mean it’s wrong it’s not  

10:00that it’s just that it’s not useful and why  genetic diagnosis is so important you know  

10:08way better than me that knowing the true cause  of the disorder ends the diagnostic policy which  

10:14is a source of stress psychological and economical  stress for patients and their families knowing the  

10:22genetic cause of the disorder allows for better  better counseling genetic concentration management  

10:29and lately in some cases you can lift  to definitely solutions and treatments  

10:36knowing what is happening genetically allows you  to access support networks such as the singapore  

10:43research fund and also to participate in clinical  and genetic data aggregations that allowed  

10:51that that allows scientific studies to be to  be developed to understand better the disease  

10:58the power comes in numbers naturally these  benefits from genetic diagnostic diagnosis  

11:05have a signage synergic effect the access to  support networks improves clinical genetic  

11:14data aggregation which in turn allows for more  pathogenic variation to be identified allowing  

11:22more patients and families accessing a diagnosis  and also better counseling and patient management  

11:29so it’s really important to access genetic  diagnosis to reach a genetic diagnosis  

11:35unfortunately for most of thing of one variance  we have one-third that it will fall within  

11:42the variance of uncertain significance realm  and we can see that it’s the most populated  

11:50category between clinical interpretation then  if we move to the missing square inside the  

11:56most common type of variants the variance of  our non-significant rates goes up more than  

12:03half of the variance being diagnosed with  this state and personal communication with

12:13is i have heard that for each photogenic single  vine that’s diagnosed choosing a variants  

12:21are identified but not all is bad  news right booze can be reinterpreted

12:30and in this matter what can we do when a  warrant is pathogenic or benign how we can  

12:36put our effort in our scientific effort to try  to improve this procedure of interpretation  

12:44what do you think what is an argument to  say that my variance is pathogenic or not  

12:50there are several criteria that can emerge  for example has my parent been tested in a  

12:56laboratory in a molecular setup do i know that  the variant caused a change a deleterious change  

13:02in the protein function is my variant  established in the general population  

13:08with no known effect has my variant been  observing other patients these are all  

13:14different arguments that arise when you sit on  the table and discuss about varying interpretation  Variant Interpretation

13:20so how to interpret the variant  and its relationship with disease   turns out that it’s quite similar to how to  solve a crime if we consider the case mr barry  

13:33mrs station we have a variant interpretation  setup we have a variant that is a suspect  

13:41accused of being causing the disease of the  victim we have evidence and we have alibis  

13:47of course the crime is the disease itself  and the judge would be the interpreter  

13:52the interpreter which can be any of us of  course within this trial we have some rules  

14:00the penal code and this was published in 2015  by the american college of medical genetics  

14:08it was elaborated in agreement by clinicians  scientists genetic counselors and laboratories  

14:14it turned it consists of 31 criteria for viral  interpretation that can be aggregated together to  

14:22make a genetic diagnosis these are the criteria  i don’t intend to show you each one of them but  

14:30it’s very important to know that there are  some criterias for pathogenicity that means  

14:35the evidence and there are um criterias for the  neutrality of the virus which are the alibis in  

14:42this case we have strong arguments for for alibis  we have supporting our arguments for alibis  

14:49not all criteria have the same strength we  have very strong arguments strong arguments  

14:56moderate arguments for pathogenicity and  just supportive arguments for pathogenicity  

15:01together they can help us to make a call and  make a judgment about the parenting question

15:09some examples of strong evidence a  well-established in vitro or in vivo functional   studies supporting supportive of damaging effect  of the gene or aging product moderate evidence  

15:22that means in a lower scale of interpretation is  that my variant is located in a mutational hotspot  

15:28and or a critical and when established functional  domain an active site opens in for example without  

15:36denying variation that’s also really important  then on a supporting level evidence that means  

15:44in a lower scale of confidence we have  multiple lines of computational evidence   here informatic methods available are plenty and  i will show you a couple of examples later later  

15:56on then we have alabas as i already mentioned if  a variant is observed with a high frequency in  

16:05the general population it’s unlikely to cause a  severe syndrome so such such as a developmental  

16:11epileptic and cellular encephalopathy normatives  here the name of the database that aggregates  

16:18variation from the general population  it’s not the only one there are several  

16:24then we have supporting alibis in frame deletions  or in search insertions in a region known to be  

16:30repetitive in different configurations and we  don’t without any known function this is also an  

16:39argument of supporting allies so having  all these criteria and arguments in place  

16:47is my very guilty or innocent  naturally we have two states   a pathogenic state a guilty state and then a  benign code and my variance is innocent but  

17:00very interpretation is a bit more complex it’s  not binary we have five different levels of  

17:07clinical significance from benign we have likely  benign then we have uncertain significance in  

17:14the middle when i cannot make a call for each  side then lightly pathogenic and pathogenic

17:24how is the very the very code well the acmg  suggests to make four different criteria  

17:30combinations that allow us to make a call when  one store one very strong argument alongside a  

17:36strong argument will allow us to make a pathogenic  call then only in the absence of benign arguments  

17:45two moderate arguments alongside a strong a very  strong argument will allow us to make the same  

17:51call of identity on the other side if i find  that this one supporting one or two supporting  

18:00arguments for benign i can make a like united code  also in the absence of photogenicity arguments  

18:10in any case that these configurations are are  not possible to make we will be obliged to make  

18:18a variance of that known significance result and  that is all other cases will be called as unknown  

18:26significance which explains also why this category  is so popular is it’s so populated in patient  

18:34repositories and clinical laboratories some  considerations about the interpretation different  Interpretation Considerations

18:42interpreters might reach different conclusions or  even oppose conclusions the information required  

18:48for each criteria is difficult to access making  a molecular history a functional experiment on  

18:56one variant it’s usually really expensive and not  many people can do this then information available  

19:04evolves constantly which makes reinterpretation  possible and it’s very important to keep in mind  

19:11that these are guidelines not rules so it’s  up to the judge at the end to make the call  

19:20i’m coming back to all the 31  criteria i would like to highlight   the computational methods available here  which leads me to the complementary methodsComputational Methods

19:34bioinformatics criteria are  considered among the acmg guidelines   there are some strong computational

19:42criteria that say if i find my variant  in patient variant repositories   i can match my variant on my hand and see if it  has been observed in other patients then we have

19:57other computational methods related  frequency to mutational hotspot  

20:02and then multiple scoring that  i’m going to describe to you now

20:09basically the information methods can be separated  into two categories the first one measures  

20:18biological context that means measuring the effect  on the chemical exchange observed if an amino acid  

20:25changes with another that is similar is not very  damaging if it changes completely in structure  

20:32and we have structural damage conservation is  my is my variant calling within the site that  

20:38it’s conserved between species this is this tell  us that the site is important if it is conserved  

20:45we also have a meta scores that combines different  scores and thresholds to generate a single score  

20:51of pathogenicity and also machine learning  associated methods here in each of these styles  

20:57i can i show you some popular bioinformatics  methods to score varying pathogenicity  

21:03that’s just one category the other category is  frequency evaluation when i use bioinformatics  

21:11databases to carry my variant and see how  it presents itself in the general population  

21:18and also in patient variance repository as i  already mentioned both of these categories come  

21:24together in the variant interpretation process  and we can use several multiple sources to analyse  

21:31in the laboratory and as part of my research  alongside with dr lao’s lab we have developed  

21:38different also some specific methods for varying  interpretation but also we lately realized that  

21:47this is not only about generating fancy scoring  pathogenicity methods we also need to understand  

21:54what’s happening and make the information  available to a broader audience and for for that  

22:02matter i wanted to also highlight simple climber  which is a method to reanalyze and and see how  

22:09patient parents fall within the gene and put the  information contained in an in a large database  

22:16in a yeast-to-use context this is how patient  variants look on timber we usually get access  Simple Climber

22:24to a big table with a lot of information that  is connected to multiple other databases but if  

22:29you’re not close to the bioinformatic or genetic  fields this table is difficult to digest i mean  

22:37the hgbs format of varying coding you usually  get the the result from a clinical genetic  

22:44test company it’s difficult to understand it’s  not straightforward so for this we developed  

22:52a database called simple climber which i invite  you to explore especially for singer and the  

22:58the tools grabs the entire database climber and um  lets you ask more questions that the ones that are  

23:07possible to make in the original database it’s the  same information but make it in a in a more easy  

23:14to digest way if you type clipboard here if you  type single you will see um a summary statistic  

23:23for the variance that all within the gene you can  automatically map the the variants that are known  

23:31to be associated with Syngap related disorders  we can see here there are a lot of them reported  

23:37the latest release already contains 680 variants  from SYNGAP1 and it’s um something that you you  

23:47can use to explore to compare your your own  variants and see if there are some trends there  Resources

23:54of course the the tools that we have developed  are not the only ones we have a lot of resources  

24:01available in the internet i can i can share  with you internally all of these sites  

24:06and there are a lot of lot of  resources lately to analyze variant  

24:12so this is a world that is on  development so we have a lot of  

24:17resources to be integrated here especially just in  the last couple of weeks we already had a report  

24:23of all the protein structures available solved  by this alpha fold computer super computer so a  

24:32lot of studies can be performed right now that may  aid in different in the future in the near futureExample

24:42so now i’m going to show you an example  of how to interpret a single variant and  

24:48again this is just an example idealist  case but coming back to the example test  

24:55we have a variant of unknown significance  here and a clinical interpretation   provided by the company we have the variant here  in the https format that i was telling you aboutInterval Tool

25:11and then i would suggest to use the judge  assistant which is called the interval tool  

25:18and it’s designed to aid in the  variant interpretation process   this is an online tool developed by the  oneglab it’s easy to use and makes all the  

25:2631 criteria semi-automatic so you can fetch  all the databases available in the internet  

25:33in just one place and evaluate how the 31  acmg criteria behaves on a given environment  

25:40we can input our variant from our own genetic test  here and see what happens at the beginning we get  

25:51a replication of the clinical interpretation  but then if we go to details and adjust  

25:57we can um obtain all the criteria and see what  we can modify according to our own information

26:07here for example after studying the variant  we realized that first this variable within  

26:13a within a gene that has a low frequency  of missing variation as we can see here on  

26:18the nomad database in the whole database  we can expect nearly 800 varying misses

26:28but we only observed 351 this tell us something  that the gene doesn’t tolerate variation in the  

26:34general population which means that  it’s likely to be associated with six   this is a supporting argument according to  acmg criteria so we can check that out then  

26:46we know that multiple lines of computational  evidence supported by protogenicity i have  

26:52selected here some popular scores that  point into the direction of pathogenicity  

27:01these are not the best scores these are  the ones that we usually use and they

27:09they work really well in neurodevelopmental  disorders but we can also check this out  

27:16because multiple already multiple computational  approaches are telling us that the environment  

27:22might be auto unit so with these two  arguments we can reinterpret the variant  

27:27and we will obtain a likely pathogenic variation  the same variation obviously but with different  

27:32criteria included here because i as a judge have  decided to include these former two arguments  

27:40but this is not the only thing that we have in  place we can we can study the parents this is  Experimental Models

27:49available and we can determine the inheritance  mode of the parent if it is absent from both  

27:56parents we can conclude that the variant is the  noble and being the noble is a strong argument for  

28:01photogenicity so we can check that out if we have  the resources and the know-how we can also make  

28:09experimental models of my variants in vitro or in  vivo to support the deleterious effect of my wire  

28:16i have access to these or two repository  repositories of experimental results i can  

28:25check this out too so we have two strong arguments  if we uh modify this on the platform we will  

28:32get a pathogenic interpretation of the variant  this time with more arguments of pathogenicitySummary

28:40then no more question your  my variant is pathogenic

28:46as a summary of the varying interpretation  process i would like to highlight that the   interpretation is a continuous process it  does not end once i make the call because  

28:56new information is always appearing it’s  bidirectional if i find a likely pathogenic  

29:03variant today it might be a boost in  the future because new information   arises so it’s highly recommended to not consider  the call the initial call and especially the  

29:17the genetic test called uh i know called  the clinical significance of my variant here  

29:23keeping up with the literature is crucial  of course each case is different and  

29:29it’s highly recommended to be conservative  innocent innocent until proven guilty right  

29:36and this is really important because  taking light in the pathogenicity code  

29:44has some strong effects it helps on first on the  on the environment because if i make a call on  

29:52a variant that is pathogenic it will stay there  and it will use as it will be used as a reference  

29:58for future goals so it’s really important to  be conservative and be entirely sure of the  

30:06code that we’re making if not then it’s better  to make a variance of a non-significant scope  

30:14of course the interpretation is better in  a multidisciplinary setup in collaboration  

30:19interpretation requires clinical molecular and  genetic expertise it is not a standalone process  

30:28and then also i i would like to highlight  that the adequate interpretation  

30:33allows also accurate candidate selection for  clinical trials and we all know that this

30:42is gaining importance as new therapies are  being developed so as concluding remarks  Conclusions

30:51for this talk i would like to talk  to you to where this field is going

30:58we all know the complexity of neurodevelopmental  disorders and until now i have only talked to  

31:03you about two states of a possible variant  a brand can be pathogenic or benign however  

31:09all of these syndromes with different  severity and stages and comparabilities   are all caused by pathogenic virus so we  need to go beyond the pathogenicity code  

31:21some approach seeing that one is  one good example of the complexity   of neurodevelopmental disorders and pathogenic  variants is not enough we need to know more  

31:31we need to know how these comorbidities are  related how how the syndrome evolves within time  

31:39and genetics can can help in the study of this  first beyond photogenic we can predict functional  Predicting phenotypes

31:47consequences bioinformatically by training machine  learning methods this is an example of a study  

31:53that i have the privilege to participate  to the story of henrik where she developed  

32:00predicting a prediction model for gain or loss  of function for sodium and calcium channels  

32:06when this has very strong effects on the  therapeutics and treatment that this patient  

32:13can have so it’s really important to know  the functional effect of pathogen parents

32:20then phenotype prediction some variants may lead  to different phenotypes severe phenotypes mild  

32:27severe mild penetrates this is seen in the syn19  when when we can have severe drug syndrome or mild  

32:38epilepsy by studying more than a thousand  patients we were able to model the pathogenicity  

32:44of variants and developed a prediction model that  was able to tell us the percentage of developing  

32:50drug syndrome or guest plus and this is achieved  just by data aggregation and statistical methods  

32:59we created a tool to provide any person  with a human science variant to develop a  

33:06prediction for their own purposes and the other  really important issue here is data aggregation  Data aggregation and diversity

33:15and diversity i show you here the map of the  epi-25 consortia the biggest genetic collaborative  

33:23study epilepsy and we can see here the  the source of patients that are studying  

33:29currently more than 25 000 but we can see  here a clear gap that i’m trying to reach but  

33:39the the study of underrepresented populations in  genetics may shed up some shed some light into  

33:45the genetic mechanisms of the whole disease  and also to uncover new mechanisms so it’s  

33:50really important and we’re all part of this  puzzle so finally i would like to end with the  Conclusion

33:58mic uh words that pretty much summary what i  would like to live with you don’t stop when  

34:07you see a bus when you have a boost in your  hands keep testing and keep asking asking

34:14with that i would like to say thank you to all  the people that i worked with especially to denis  

34:21lal and also to lisa marie who developed this  nice allegory of [ __ ] of judgment the trial

34:30thank you that was great i think we have  uh questions i have a few questions too but  

34:36Victor is asking if you can see  the journey what does he mean  

34:42when a barrion is determined to be  a boss based on a law problem scoreQuestion

34:51what does it mean for a variant to be yes

34:56do you see the Q&A questionnaire  are you going out here

35:08i’m sorry i do not know  what it’s probably and score   i haven’t heard of this if somebody can help  us out in the chat with the province score  

35:20i don’t know jay i don’t know what is  probably honest for i don’t know either  

35:25yeah but none of the scores uh especially the  bioinformatics ones are definitive scores uh  

35:35scores can be integrated to the acmg guidelines  as computational evidence but they do not have  

35:43the power to determine pathogenicity  by themselves that would be my comment  

35:49for any score okay and i have a question how often  um i know we talked before that initially when the  

36:02companies started doing the genomic testing there  was a lot of positive results and i want you to  

36:07explain all that to people and when is the cut off  when everybody was like a wait wait you are doing  

36:13too many diagnosis and why that happened i want  you to give a perspective what happened to people  

36:22if if you take a look at the rates  of pathogenicity delivered by   genetic test companies at the beginning  it was high and mostly because there was a  

36:33positive pressure to establish clinical genetic  testing in as a field as something that it’s  

36:41useful this is this i don’t think this  was intentional it’s just how it happened  

36:48then as we went in to increase the volume  of variance and clinical variants analyzed  

36:55people started to be more conservative because  initial pathogenic calls were not that pathogenic  

37:03when you started to analyze more patients  and that is one of the reasons that all  

37:09the stakeholders gather together and  establish these acmg guidelines to  

37:15make the process systematic and most companies  today they try to get close or even beyond  

37:23the guidelines today to to be  more precise in this process

37:30okay and then um the other question is uh how  often you get second opinions of i mean how how  

37:38that works because we know we have patients that  come to our using god saying you know how i have  

37:46a VUS for Syngap the kids look similar to our sync  ups but it’s still they are in a VUS diagnosis how  

37:54can you regular way or to recommend to railway for  boss and how do you ask for a second opinion to a  

38:01higher institution like yours that have somebody  who work in this very hard second opinions  

38:09um well i i think one of the most important  things is to have access to genetic counseling  

38:18after a genetic test because most of the  processes done by clinical testing are automatic  

38:25i know they provide most of the times  a report evaluated by a scientist  

38:32but still the genetic test is fixed  in time it cannot update itself  

38:39because it’s it’s paper right and you may  stay with that and i do not recommend that  

38:44i would recommend to revisit the interpretation  um i don’t know once a year for viewers u.s uh  

38:53information is growing and growing exponentially  so reports are getting better and better

39:02so and and also there are the tools to study   and i’ve seen how experts and how fast  parents can become experts in the field and

39:16so you are able to do your own interpretation  with all of these tools that are available  

39:22and the community is quite  active you you can always ask   some scientists that it’s researching the gene who  have the molecular expertise to make some calls to  

39:33modify some criteria and i also recommend that  to reach out to scientists who most of the  

39:41times love to help and love to contribute  to things that are quite important right

39:51have you seen diagnosis that you have  revered from pathologic to benign  

39:57or possible benign have you done that i i  haven’t done it myself but i’ve seen it it’s  

40:05mostly because when these general population  databases they have like a new release and  

40:12they report like twice the size you start to see  variance that wasn’t in previous database but  

40:19now you you see them and that’s a strong argument  for perpendicular variants so that has to happen  

40:28okay do you have oh i think we have more Victor  and jazz are asking more of the q a and i think  

40:38yeah JR may have some more important questions

40:43yeah well i wanted to talk a little bit more  about victor’s second question so he had  

40:48said that so provine is one of the i think  it’s uh at the ventner institute it’s a  

40:56it’s just one of the tools you can use to look  for how how different you think the amino acid is  

41:02basically um so it seems like for his questioning  the idea is to go back maybe once a year  

41:11and continue looking at the can continue inquiring  about the VUS with the geneticist or the genetic  

41:17counselor yeah um so a year sounds like a long  ways away but we can also do some of our own  

41:24research on it and try to bring this information  to them i also think make sure we’re in citizen  

41:30you know because they have a really top-notch  genetic counselor and ellie brimble and she  

41:36as far as i know she personally looks at each  variant and uh tries to figure out what’s going  

41:44on so i guess that’s i guess that’s maybe the  opposite because you’re supposed to know that you   have a pathogenic to go to citizen but i think my  question for eduardo which might might have to do  

41:56with victor’s question i’m not sure is how much  weight is given when you have a small number of  

42:04genes on the test so say you’re doing an epilepsy  test an epilepsy panel with just 151 genes on it  

42:12is there is the same weight given to  that as when you do an entire exome

42:19right because because you’re only if you’re only  looking at 151 and you have a vus what about the   rest of the genome that wasn’t looked at how i’m  just wondering how does does that have interplay  

42:30in how these are scored or no well well um gene  panels evolved constantly the literature reports a  

42:42new lfc gene every month if not so so these finals  are constantly changing and it’s a source of  

42:51delay in the diagnosis process because when we  have the exon the complete action we can do two  

42:58things we first we can interrogate all genes  possible and then we can bring revisit our exon  

43:06with when time passes and see hey we have a couple  of genes reported has my baron been there so exams  

43:16tend to be uh of use for a longer time than  panels of course the panels they have were the  

43:27a relatively high chance in a cost cost effective  strategy and what i’ve seen also is that patients  

43:36who have a negative panel they go then to  exome sequencing and this introduces a delay  

43:45in the diagnosis with all the effects that  this has so it’s really important to achieve  

43:51diagnosis fast and for that i will always  recommend if you have the access to an axon exon  

43:59does not provide all the answers and there are  several ways of identifying pathogenicity that are  

44:05outside the axons and it’s um it’s part of the  strategy um i i’ve seen so so you’re talking about  

44:16also uh copy number variation and carry typing  other tests that have other and even i guess  

44:24uh mitochondrial testing yeah yes poison actions  structural variants repetitions uh short tandem  

44:32repeats all of these type of variants uh are  not that easy to see on panels or exons so um is  

44:42if you have a uas on the axon it’s  does not mean that you don’t have  

44:47that you don’t have a photogenic variant  there are other tests that can determine   alternative mechanisms that could be  happening but we are just not seeing them

44:59yeah i have other questions but i think  maybe oh i saw victor and jess are both  

45:05here marta do you want to have them talk  first i’ll ask i’ll ask when they’re done   yeah victor do you want to ask your question  sure thank you very much for the talk it was  

45:16really really helpful and um i’m a parent my  son was diagnosed with sin gap 1 based on a  

45:23vus but with like the his his medical history and  all in his the way he presents he was diagnosed  

45:31um so this has been super helpful it’s i only  understood a little bit of your talk i think but  

45:37um oh i’m sorry no no it’s okay uh i i’ve been  trying to learn it on my own and uh it’s it’s  

45:45definitely a complicated world um and i actually  work in medicine too so it’s a little i have some  

45:50of a head start but um my question was around  like where should we focus as like a patient  

45:56group in terms of helping advance the science of  classifying variants i i know we actually enrolled  

46:04we’re enrolled in like this all of us which is  like a us-based population um cohort where you  

46:11contribute your your blood samples and they do  genotyping i think whole exome sequencing and then  

46:17you also contribute all of your your electronic  health record data and surveys do you see those as  

46:23being helpful for this or is it really you should  we should be doing things like the citizen and  

46:28things that are like really like more focused on  like rare diseases i guess and what’s most helpful

46:37well um there is a limit in what you can do on  a genome-wide level to determine pathogenicity  

46:46not all genes are the same not our dna  segments behave the same so i think  

46:54moving into the gene wise direction uh is the it’s  uh it’s a good strategy and and for that uh data  

47:03aggregation is key we we need and to accumulate  data to extract meaningful information um  

47:13the more data and patience we have the more the  more follow-up we do we can understand better how  

47:20these variants behave in individuals in multiple  individuals over time so as a group i think  

47:30the a good strategy is to is to accumulate  numbers and data um also also to develop more  

47:39more robust statistical models  specific to to certain phenotypes  

47:46and for us that’s by joining citizen right yeah  that’s the major work we are trying to do now  

47:54is gonna be phenotyping all the uh genetic  uh data that we have that’s that’s the  

48:00short-term goal the long-term is different but  the turn-turn goal is to get the phenotyping  

48:06of all those variants and group them that you  are doing a lot of that yeah is doing a lot of  

48:12that that that work grouping and everything but  yeah we are general that’s that’s the goal the  

48:20are you able to enroll if  you have a vus into citizen   uh they we if you have a us we have i think  uh they are like more than 10 by now and

48:36victor said he had a diagnosis though he had  a diagnosis plus the genetic vos so that’s uh  

48:42yes you should enroll if you have a diagnosis of  singaporean you should enroll for 100 sure okay  

48:48if i’m if i’m able to complement my my previous  answer data aggregation is really important and  

48:54also molecular studies the the molecular  characterization of genetic variants found  

49:02in patients i think this is key and also  every functional characterization stays  

49:11as a as a value of information for  the future so it might be expensive to  

49:17characterize uh every possible variant  but every little step that we make in  

49:23that direction will help to uh get the full  picture of variability in singapore for example

49:33so can i ask you to elaborate on  the molecular characterization   are you talking about like biochemist biochemical  assays in vitro you’re talking about mouse models  

49:44yes any any any kind of characterization  whether it’s in vitro and diva okay yes  

49:52yes and any kind and for singap we were discussing  before the talk that it’s quite complex because of  

49:59its function interactions location it’s  quite complex but every insight that we can  

50:07make into the direction is valuable  and it will be it will serve the future

50:17yeah thank you so just had a question martin  yeah go ahead yes hi can you hear me hello hello  

50:28um so my question is uh with the reinterpretation  of the bus after a period of time um  

50:37from my experience it really relies on the  families pushing for the re-interpretation   in many cases and i’m wondering if there is  are we converging on a generalized approach  

50:48to automatically re-evaluating um experience of  unknown significance in the light of new evidence

50:58i know that some companies  they provide reinterpretation   amongst certain conditions um make it automatic  uh it’s i’m not sure if that is possible because  

51:13the nature of the new evidence is it comes  from so many sources that it requires um  

51:22the it requires on the table different people  rather than a an automatic method um and on this  

51:29table uh clinical expertise molecular expertise  bioinformatic or genetic expertise are required

51:40so i i don’t think it can become completely  automatic in the near future thank you so  

51:53uh so as a follow-up to justice question you  have you showed us a great website for looking  

51:59at dravet that i think you guys run right is that  from your lab yeah yeah yes we developed that with  

52:08its own and revision currently so are are  you developing the same thing for some gap  

52:14one or are you developing the same thing for all  genes or is that something that is a would need  

52:20to be a funded project to do foreign we’re not  developing something in those lines for symbol  

52:27sync at one we are working on  related sodium channels because it’s  

52:34we have figured out that it’s easy to study them  together so we’re working in that direction with  

52:39scn2a and another sodium channels but not on  zynga currently it will be possible with the with  

52:50data available and that method was developed  based on data aggregation from multiple patients  

52:59variants age of onset clinical features like that  if we wanted to try to set up a project like that  

53:06how would we go about that as a patient  advocacy group how would we figure out who   should do that same kind of project for  some gap that you’re doing for dravet  

53:16right because we don’t want it we want to know we  want to find the actual syngap patients we want to   find the people that don’t have syngap you know if  they think they do we want to know what the which  

53:27parts of the gene do what and you know we want  to know it all so it seems like this type of um  

53:35this type of it’s it’s almost like a  hunting you know gathering expedition   to to find all this data from all these  disparate sources and and interpret it

53:49well and uh we started the scm-1  project with one question can can we  

53:57differentiate a mild presentation of the disease  with severe and if we can do that we think of  

54:07which i think we could on based on the data  aggregation that you have might be possible

54:17okay so maybe you have a postdoc that  might want to just sure we could okay  

54:24okay we could talk about that more i i have a i  don’t want to put you totally on the spot about   that but we this is something we all need like  our our patient community desperately needs us  

54:33and we don’t we’re all kind of feeling around in  the dark and i’m i’m you know maybe the leader of  

54:39feeling around in the dark but i can’t you know  people are like oh i don’t you know i i can’t   do this this isn’t something i can do but it’s  something that we all think about and that we need  

54:47we need um concerted you know we need a concerted  effort toward it so it’s very exciting to think  

54:54um okay so i have a i want to ask you a  question in a different realm if that’s okay   um you talked when you showed us your the  different types of mutations one type of  

55:05mutation that wasn’t on your chart was uh sort  of upstream very you know variations in the  

55:11upstream regulatory regions so something  that might change the expression level  

55:17and i was wondering if you either for dravet or  for sync gap if you look at upstream variation  

55:27and um there’s there’s two reasons one is  to want to know about how expression levels  

55:34because if half is so severe maybe you know  maybe less is is going to be some sort of mild  

55:40right some sort of mild phenotype and then the  second reason is as we get toward precise genetic  

55:47therapies there are some times when we might  want to distinguish between the two alleles   and so knowing if there’s general variation in the  population already that you could we is it likely  

55:59that we’d be able to distinguish between the two  alleles just by um normal population variation  

56:07do you see what i’m saying not that we know  a priori which it was but you know you might  

56:12have to figure out which which one is which  but is there a variation there that would  

56:18allow us to distinguish between  the two alleles in most humans

56:25well the the question has has several branches and  it’s it’s complicated sometimes to differentiate  

56:34that something is not there because it’s not there  or because i’m using the wrong glasses let’s say  

56:41or the wrong platform for utr’s structural  variation repetitions sometimes there is a  

56:49technical difficulty in seeing those variants  to increase our own awareness that maybe we  

56:58have some singaporeans that are increasing  the expression on a non-coding region and  

57:05non-coding regions are not tested in most  of the clinical genetic testing approaches  

57:13so that’s a technical uh difficulty if we have  the resources we will test the whole genome and  

57:21and we will explore everything and even that  will not be enough to detect all pathogenic  

57:26events and that we have we know that there  are at the transcriptional level some some  

57:33errors that can may lead to gain or  loss of function of certain genes  

57:39so it’s really important to be  aware of the methods that we have um  

57:47unfortunately is once the test is done it’s not  possible to revise things that are not red so it  

57:55might be difficult to evaluate that okay so in the  in the epilepsy gene paddles they don’t look at  

58:02some of the upstream because i thought in whole  exome they lived at least a little bit outside   the exo sure excellence sure in the collection  we have uh we have some proofs that detect utr  

58:17you can you can you can even design your own your  own proofs to scan those regions more in deep  

58:25but it’s we’re still learning how how genes are  expressed so that’s also uh that’s also something  

58:34that we need to consider the poison exons are  a good example they are difficult to detect in  

58:42in exome sequencing because you don’t  have the proofs because you weren’t aware   that this portion of the genomes were  actually axons that appear suddenly and  

58:54so we also have regulatory  regions that are far away from  

58:59from the gene itself so these regions in  trance that regulate expressions and some eqgls  

59:08it’s it’s not entirely reduced  to the to the segment of the gene

59:19okay well i’ve of course asked something  that can’t be answered so sorry about   that um no no no it’s it’s an important  discussion to discuss different mechanisms

59:34yeah i was part of a discussion with some  people from some ring chromosome diseases  

59:39and they were saying that their their diagnosis  rate is going down because people don’t do carry  

59:46type analysis as much anymore so as as all as  you know things being diagnosed with panels and  

59:53and whole exome are going up because of the  increased testing uh their diseases are are being  

1:00:00less diagnosed so i found that interesting that  the right the things that we’ve known for a long  

1:00:07time are starting to not be diagnosed as much just  because the tests are falling out of favor um okay  

1:00:13so it looks like mike would like to answer some  questions live so i’m going to stop talking now

1:00:24i think that i think we answered all the ones  that were in the q a but yeah i think all that  

1:00:31you guys were talking about it goes to a relevance  that uh for probably uh people that has bosses  

1:00:38but uh going to the i mean genetic counselor once  a year probably is very valuable because that’s  

1:00:46the problem uh i mean this is a science that  is just beginning this is it’s a lot to know  

1:00:54and a lot of things that probably will come up  in one year two years three years five years  

1:01:00and for the ones we that we have the pathogenic  diagnosis we probably will be a smart at least  

1:01:07every three years i think what you think i think  it revisited that diagnosis is probably important

1:01:16yes for sure for sure um ues are um

1:01:23must be so frustrating for for people  having having that test in their hands and

1:01:32it’s a it’s a responsible claim to say that  reinterpretation might give you hope uh if not now  

1:01:42maybe in a year or so we might as the science  community accumulate enough information to make  

1:01:48the call yeah another thing is for example i have  a sophia my daughter has um the the pathogenic for  

1:01:58Syngap but also has two VUS in the in her genetic  report then there is something now that is coming  

1:02:06to evidence in the genetic that everybody has  a different background too then how they will  

1:02:13show the symptoms also the pain of the genetic  background as a whole right yeah hey that’s uh  

1:02:22that’s also really true because you have  causing factors but you also have modifier  

1:02:27factors that are maybe not entirely related to  single one but just the background as you say  

1:02:36for that matter i’m also really interested in  gathering information from underrepresented   population or also people with mixed genetics  like admixture that it’s so common in latin  

1:02:49american populations that background uh might  explain why some patients uh react differently to  

1:02:57drugs or manifest different this is progression  um or other rare variants within them so

1:03:08yeah that’s really really true   okay i guess any any other questions i think  we are at the end of the hour any questions or  

1:03:19it was great thank you so much Eduardo and this  is i think very clear and very helpful for a lot  

1:03:27of families thank you and i hope that santiago  gets a little bit warmer too thank you for that  

1:03:36anymore and we truly appreciate very much you are  doing this for us yeah thank you stay connected

1:03:46we weren’t talking before about working together  and we are still thinking about it yeah we are  

1:03:52trying to do all these projects that are  very important and of course your project   is very important to us thank you so much thank  you so much it’s really wonderful thank you