59 – Finding more SYNGAP1 patients with Probably Genetic

Take the Probably Genetic Syngap survey here.

View our Probably Genetic news release here.

These are our introductory comments: 

I have the pleasure to introduce today’s speaker, Abby D’Cruz, who is Head of Strategy & Operations at Probably Genetic.  Probably Genetic develops technology to find undiagnosed patients
online and helps get them tested. Abby is an expert in genomics, holds a PhD from the University of Oxford and specializes in rare genetic diseases. Before joining Probably Genetic, Abby was a Senior Consultant at Trinity Life Sciences.


0:06i have the pleasure to introduce  today’s speaker abby de cruz who   is head of strategy and operations at probably  genetic probably genetic develops technology  

0:15to find undiagnosed patients online and helps  them get tested abby is an expert in genomics  

0:22holds a phd from the university of oxford  and specializes in rare genetic diseases  

0:28probably genetic develops technology to find  undiagnosed patients online and helps them get  

0:33tested before joining probably genetic abby was a  senior consultant at trinity life sciences welcome  

0:40abby tell us a little bit more about probably  genetic and what drove you to join the company  

0:46hi everyone thanks very much for having me really  excited to be here and with the team at the SRF   as well so thanks Corey for the introduction  uh and it’s a great question because i’ve been  

0:57with probably genetic uh coming up to i think  four or five months now and as you mentioned  

1:02my background uh after my phd in uh which  kind of was oriented around clinical genetics  

1:08very much in a lab research focused position after  that stint i i joined trinity life sciences where  

1:15uh you know i really had an exposure to a whole  variety of therapies and also disease areas but  

1:22i developed a sort of a focus on rare diseases  orphan conditions and then also therapies that  

1:28were being developed across those indications so  through that position i think got a really broad  

1:34exposure to some of the challenges that uh drug  development faces uh within those settings and  

1:40i think through probably genetic i’ve been able  to really zero in on uh the most i think pressing  

1:47challenges uh that occur in this space and some  of the most kind of incredible and innovative work  

1:52that’s going on to address those challenges  because you know the mission as you just   mentioned Corey at probably genetic we have a very  aspirational mission to diagnose uh 200 million  

2:03uh genetic conditions uh to diagnose 200 million  patients with the rare genetic conditions globally  

2:09and you know that’s obviously a pretty impressive  target but we’re doing that through really   innovative ways that i think are really exciting  and we’re kind of looking to leverage some very  

2:20sort of cutting edge and and you know newly  developed ideas that merge technology uh you know  

2:26in addition to you know all of the ridiculously  advanced research that’s currently going on  

2:31in the genetic field so looking to merge a couple  of my areas of interest and tackling challenges  

2:36that i’ve come to understand are really complex  and obviously you know incredibly worthwhile so  

2:42we managed to take quite a few boxes at probably  genetic and the work is incredibly rewarding as   i think we’re going to get into today wonderful  look forward to it perfect well i think everyone  

2:54can see the screen that we’ve put up today  so we’ve got a short uh presentation where  

3:00we can go into really some you know background  on what it was that we put together with the SRF  

3:07and uh you know what the kind of objectives were  for this partnership and and excitingly what’s  

3:12kind of come out of that program and uh looking  ahead to the future of next steps as well so  

3:17to kick it off at a very high level our objective  was to figure out whether or not we could   identify SYNGAP1 your sort of target condition  of interest using machine learning effectively so  

3:29what we’ve broken this down into in the  next slide if it will move there we go

3:37uh to pursue that objective of developing a what  is effectively a symptom checker that we could   use to identify SYNGAP1 patient we break that  down into essentially a five-step process so  

3:48the first step was to develop a prototype which  essentially involved the probably genetic team  

3:55working really closely with the SRF as well to  leverage you know a lot of researchers out there  

4:00currently in addition to patient feedback family  feedback care feedback to develop a prototype  

4:07symptom checker so what we mean by that really  at the very initial stages was just effectively  

4:13a list of symptoms that would be associated with  the target condition so the SYNGAP1 but also a  

4:18list of symptoms that would be associated with  differential diagnoses that patients experience  

4:24in their diagnostic odyssey and then also  just kind of you know peripheral symptoms that  

4:30might crop up as well so we’re really trying to  capture a comprehensive list of symptoms that our  

4:36patients might encounter within their journey and  feed them into what we developed via our platform  

4:42was a really user-friendly symptom checker  that patients would go in and for example you  

4:48know use more patient-friendly terms for some of  these phenotypic symptoms that might crop up so  

4:53we initially developed that prototype uh and then  what we did again leveraging the this the SRF  

4:58community was that we put that out for feedback  effectively um we put that prototype out to say  

5:04are these the symptoms that you’re seeing  is this comprehensive is this accurate   um is there anything that we’ve missed really  importantly um you know there are any terms that  

5:12maybe are surprising to you um that would help  you to use to identify what’s going on with you  

5:18in a way that uh you know would enable you to  input this into a data capture form so we honed  

5:25that prototype in the feedback step and once  we were pretty confident that we had a really  

5:31comprehensive and accurate symptom checker we then  set out to collect training data uh effectively so  

5:38our target here was on 100 patients so what we  mean by collecting training data is essentially  

5:44just getting people to fill out the symptom  checker so uh sending this link out um that  

5:49was live on uh you know definitely within the SRF  community lives on your website as well we got an  

5:55enormous response but effectively we just wanted  to get that link out to as many people as possible  

6:01uh so that we can collect training  data on patients within the community   to see how those responses look when we submit  you know this huge list of symptoms to them  

6:12uh what what’s coming up right so what are the  patients within this community how are they   characterizing their symptoms what’s the what  we call a phenotypic profile for this community  

6:22so that was kind of the training data step and and  like i mentioned we’ll go into data a little bit  

6:27later but our target there was was for a hundred  patients um for a really good comprehensive  

6:33snapshot so once we’ve developed the symptom  checker we’ve received the feedback on it we’ve uh  

6:38basically unleashed it on the community uh  we’re then training it so the training step   it’s uh it’s you know encompassed in three words  here but it’s it’s a little bit more complex  

6:48than that this is where our machine learning  technology our sort of algorithm-driven technology  

6:53comes into play so what the symptom checker can  do with our machine learning algorithms is it will  

6:59ideally get really good at distinguishing  patients that have the target condition to SYNGAP1  

7:05um from other patients or patients that don’t have  signed up one so we would know that in the you  

7:10know within the community we know patients that  have previously received a diagnosis of SYNGAP1   so the goal there when we say training the symptom  checker is that we would get it to a point where  

7:19it becomes really good at identifying when someone  who’s submitted the phenotypic profile to the  

7:25symptom checker has a high likelihood or does  have SYNGAP1 as the target condition so that’s  

7:32that fourth step um and then finally you know the  the kind of the why or the so what of this whole  

7:38exercise is to basically provide um the SRF with  an incredibly rich source of you know information  

7:46uh really useful information uh for their  community and i think that the fifth step  

7:51is really obviously the really exciting  one definitely for the SRF but also for us   because um you know that’s the real value add  i think and that really enables us to zoom out  

8:01a little bit on the why um because you know as we  know and as as certainly everyone at the SRF knows  

8:07within the rare disease community you know the  more patients that you can find and the more   you can draw attention to how many people are  either living with your target condition or you  

8:16know trying to find answers um in the space  within which your kind of condition exists  

8:21the more people you can identify in that space the  more attention you can drive onto that disease the   more potential r d f that you can drive and the  whole kind of you know biotech and research cycle  

8:30can um have a bit more energy put into it so that  was a little bit of an overview of of the initial  

8:36five steps let me pause query to see if we’ve had  any feedback or questions or if i can keep going  

8:42let’s see i don’t see any questions as of yet  great i must be doing a spectacular job or they’re  

8:50all just holding up so i’ll keep going so this is  basically the overview um where we’ve broken down  

8:56into steps uh how we were kind of addressing this  overall objective to develop a symptom checker   to identify SYNGAP1 patient so let’s get into how  this actually fared so um like i mentioned just  

9:08before the initial objective was that um you  know in that training step where we were going   to collect submissions we were we were targeting  100 as a really healthy data set what we achieved  

9:18at the end of our initial pilot was 603 unique  submissions uh and that was right across the globe  

9:24certainly right across the um the united states  and you can see from uh on the left the the growth  

9:30tracker in terms of number of symptom checkers  this program uh had just phenomenal growth which  

9:36was powered um by the uh you know enormous level  of motivation and engagement driven within the srf  

9:42um for people to submit their symptoms and  and create those symptom checkers submissions   so we had uh we saw we witnessed rapid growth  of the program and the geographic spread as well  

9:53is a really powerful um i think example of just  how much appetite there is um and just how um  

10:00you know just the need i guess the high level  of need for people out there who are looking   for answers people out there who are looking for  remote testing programs and patient initiated  

10:10uh solutions so we love this slide um you  know it really illustrates the scale of um  

10:16you know what we’re doing and certainly  illustrates on how motivated the srf community is   in particular um to you know click on the link  basically and distribute the link to patients and  

10:26their families who are looking for for answers  so what we’re at now is 3287 unique submissions  

10:33and we’re seeing that continue to grow and  seeing that to continue come in from around   the world so massively exceeded our our growth  expectations in terms of program submissions

10:46so the next piece is just what those submissions  or what that kind of snapshot of the community   looks like so the people that are submitting to  this program um again this is for the current  

10:56data so i think by the end of the pilot we had  about an 80 we had 80 saying get one patient   submit to the program we’re currently at 100 sign  up one patient submitting to the program and uh 20  

11:07different diagnoses as well um that are submitting  uh phenotypes to uh you know to the program and  

11:13contributing to the richness of that data set  and contributing to the richness of information   that we have of patients that are within the SRF  community so this is a really powerful graphic as  

11:23well just because it illustrates just a variety of  you know differential diagnoses and also just the  

11:28variety of conditions that are present within this  community which again you know on this on this  

11:34graph we’re showing some incredibly rare disorders  um and again it shows uh you know what we think  

11:39kind of underlines the um diagnostic odyssey that  a lot of patients go through and again provides a  

11:44lot of really really useful information for the  srf and similar organizations that are really  

11:50looking to shed light on some of these conditions  so some pretty extraordinary numbers in here

11:58so on the next slide uh what we’re showing here  and what we think is really powerful is um you  

12:04know in terms of objective of achieving that  objective of a symptom checker that can identify   SYNGAP1 patients what we found through the program  is that SYNGAP1 has that distinct phenotypic  

12:15profile as compared to patients that don’t have  SYNGAP1. so those are the patients that haven’t   submitted you know a diagnosis of of syngap so  what we’re seeing on the left here is essentially  

12:26a uh ranked uh snapshot of the symptoms of  everyone that self-identified as having SYNGAP1  

12:33so those are 84 patients that we use there and  you can see that the global developmental delay  

12:39the intellectual disability motor delay are the  top symptoms that are incredibly frequent within  

12:45patients that have the same gut one diagnosis  whereas on the right hand side you can see that   patients that don’t have that SYNGAP1  their phenotypic profile often looks  

12:55very very different to those SYNGAP1 patients  so this is really just underlining the power  

13:01of sort of addressing the question that we  had around you know can our technology can   a very comprehensive symptom checker and our  machine learning algorithms can that you know  

13:11with a high degree of likelihood identify a sign  got one patient from a non sign up one patient  

13:17and what we found is that look the snapshot  looks pretty good this is really promising   uh this is what we like to see again we  love this slide um because it’s just sort  

13:25of illustrating the power of this program  because it is a patient-initiated program   so we’re just sort of putting this out  there in the community and these kind of  

13:33stock results is really kind of underlining the  power of the program in terms of predictiveness

13:42so we’re just going to dive a little bit into  um some of the i guess the science or the step   four if you remember of the training um for our  tech because we think that this is kind of really  

13:52powerful and it’s sort of the nuts and bolts  of the slide that you saw previously so this   is just a little bit on what our model actually  does and what it does is that it will look at  

14:03all of the users that are the SYNGAP1 submissions  all of those profiles and it looks at the symptoms  

14:09that those patients submitted and it’ll rank  those symptoms by importance so what that means is  

14:14essentially um you know if there’s a symptom that  the vast majority of patients with syngap have  

14:20that’ll be a really important symptom for  the model it’ll prioritize that symptom   and kind of flag it as something that is uh seen  very often very often with saying get one patient  

14:32and it’ll also mean that um you know the  importance of that symptom will increase   relative to non-SYNGAP1 patient so this is  basically just a quick look behind the scenes  

14:42and again really important information for  communities like the SRF where they’re looking to  

14:47really sort of have an indication of the symptoms  not only that their patients are experiencing but  

14:53also the the symptoms that really distinguish  or differentiate their target condition and can  

15:00be used to sort of identify patients that have a  high likelihood of having SYNGAP1 in this instance  

15:06so this is what we refer to as feature  importances but it’s essentially just   allocating an importance to the submitted symptoms

15:18so what we’ve called out here is just a couple  of examples where you know we you can what this   enables us to do so for example it will enable  us to identify for example a simple a symptom  

15:29that might be um common for SYNGAP1 patients  but rarer for non-SYNGAP1 patient so what that  

15:36will enable us to do is to flag that symptoms  as an as an important one because it distinct  

15:41it can potentially distinguish between someone  that has signed up one and someone that doesn’t   because it’s effectively telling us okay  this is a common symptom in saying that  

15:49one patient but it isn’t in confirmed  non-SYNGAP1 submissions this is really  

15:55powerful information when you’re looking to you  know go into that sort of characterization of a   patient community and flag those symptoms  that are really important for your target

16:03disease so the so what of  everything that i’ve kind of just  

16:10said in terms of you know the things that our  model is doing behind the scenes to identify   important symptoms and to be able to distinguish  99.1 users is basically shown on this graph so  

16:20the key message here is that through all of those  mechanisms that our machine learning model uses  

16:26um we found that we can use it to identify  ninety-seven percent of SYNGAP1 testers  

16:31so what that means is of if you had a hundred  SYNGAP1 patients that submitted to the program  

16:36the machine learning model that we trained  just by itself would be able to identify   97 uh 97 of those uh signed up one patient  so a really really effective um predictive  

16:47uh rate basically um what we did there was uh  you know what this this graph is showing is  

16:53that we identified a threshold that we used um to  basically score um submissions to say this SYNGAP1  

17:02submission uh you know is is kind of over the  threshold and therefore has a high likelihood of   having signed up line or should would therefore be  referred to testing or something like that so um  

17:11the slide here is basically showing that the ml  model uh has a really high uh ability to identify  

17:17SYNGAP1 testers which is really good news for  us um based on all of the work that we did   with the SRF training this into checker and then  having it be able to identify sign up one patient

17:33so this is kind of the uh outcome slide and  this is accurate for the end of the pilot   so um i won’t get too much into this but the  the key message here is that with the ml model  

17:44that we trained working with the SRF we had 81  syngap submissions that would be offered testing  

17:51and really importantly we had 28 uh non-SYNGAP1  submissions that we would also offer testing  

17:57so those are patients that we’ve uh that have  submitted to not having a SYNGAP1 test and not  

18:03having a SYNGAP1 diagnosis but that the program  would offer would suggest to offer testing too  

18:09which is really interesting and valuable for  us and that’s only at the end of the pilot   so you know we’ve collected a whole bunch more  submissions since then and we would anticipate  

18:18that that number will go up so this is exciting  on two levels um the first is you know that we’re  

18:23identifying um patients that you know the model  is kind of accurately identifying patients that  

18:28would be offered testing because they um you know  look like SYNGAP1 patients but really excitingly  

18:34is that it’s identifying patients that haven’t  currently got a diagnosis of SYNGAP1 that we  

18:39think should be or would be offered testing which  is great so again really valuable for um you know  

18:46organizations like the SRF who are not only  interested in identifying patients in learning  

18:51more about their community but also in identifying  patients so it was really good news for us  

18:56certainly but also this kind of data is really  exciting because it hits on both of those kind of  

19:02key objectives for patient advocacy organizations  and certainly for ones as engaged as the SRF

19:14so this is kind of uh diving in a little bit to  what uh to the previous uh phenotypic profile  

19:19slide that i showed where um if you kind  of double click so on the previous slide   we’re kind of zooming in now on the on the 28  and the 81 so of the 28 and 81 so over here  

19:30on the left hand side where we know these  are patients that have SYNGAP1 that’s their   phenotypic profile and this is just kind of  illustrating that of those 28 patients that  

19:40don’t have a diagnosis of SYNGAP1 that the  program has sort of flagged as you know should or  

19:46would offer testing for SYNGAP1 their phenotypic  profile looks much more similar to patients that  

19:53do have SYNGAP1 so this is basically you know the  foundation upon which our program suggests that we  

19:58would offer that kind of would flag these patients  that the SRF can you know connect with these  

20:04patients to link them with additional information  and resources this is the basis for that where  

20:09we’re saying hey these patients have a phenotypic  profile that we think is pretty similar to assign  

20:14gut one patient and therefore it might be helpful  that we you know connect them to resources and   ideally additional testing so this is a really  sort of exciting and obviously useful output  

20:26for programs like foundations like the SRF but  also is kind of the key reason behind programs  

20:33that we we run like this where we can hopefully  use our technology to identify patients that  

20:38haven’t yet been given a diagnosis but that um you  know could certainly be really good candidates for  

20:43additional information and potentially linking  and connecting to testing resources as well

20:52so in summary essentially we wanted to show this  slide as kind of a snapshot of at the end of the  

20:59pilot so at the end of our pilot that we put  together with the SRF we had that objective  

21:05that i spoke about earlier to develop a symptom  checker for SYNGAP1 and we broke it we had broken  

21:10that down into five steps um so the great news  was that we met or exceeded all of those steps   we developed the prototype and received feedback  with a huge amount of help with the by the SRF  

21:20and their community um we had that training data  huge overshoot 603 patients ended up submitting  

21:26so a really powerful training set that we  were able to run the symptom checker on   uh and developed what i’ve  hopefully just shown you uh  

21:34to be a really accurate and uh highly predictive  symptom checker that’s really powerful for SYNGAP1  

21:41and then we were able to provide a huge amount  of really exciting data to the SRF as well  

21:46that you know has hopefully really empowered them  and you know provided them with an additional   uh very rich resource um to continually sort  of drive engagement um and effectiveness within  

21:57their community and we had sort of put in as kind  of aspirational bonuses um i remember at the end  

22:03of our sort of pilot report with the SRF that um  we had two additional points that we were really  

22:09excited to get going on which was firstly to  build a tool for the SRF to engage you know your  

22:14community and your patients where we could put all  of this data and all of these abilities together   into a really user-friendly portal and then  the second piece which is kind of the most  

22:24important bit was to launch a free testing  program so that we can basically connect  

22:29these patients that we’re flagging as you know  potential candidates for SYNGAP1 we can connect  

22:35them to that potential answer in the form of  free testing so this is what we kind of showed  

22:41at the end of the pilot that we put together  with the srf and we were really delighted to   have met all of the uh the project objectives  but the exciting uh update uh and i don’t even  

22:50know that the srf knows this at the moment  but the excited update that we have to share   today is that uh we actually have just signed  a partner for the uh free testing program and  

23:01uh we’ve just signed with uh mighty therapeutics  who are gonna basically work with us uh to  

23:08essentially be able to offer um potentially offer  free testing to any of the srf submissions that  

23:13are eligible so we have uh that capability kind of  the back end to this uh in process which is really  

23:21really exciting and kind of connects up the uh you  know the finding with the with the additional sort  

23:26of diagnosis piece as well so that’s a piece  that we’re really excited to get working on um   and and was our sort of eventual goal so that’s a  really exciting next step that uh we have to come

23:39wonderful i know everybody is really  exciting about the fantastic program

23:47um what else let’s see we’ve got a couple  questions let’s see if i can get them pulled up

23:58before you before you jump into that  cory can you um i have a question abby so  

24:05on bullet three there you say collect train  data on 100 600 3 000 200 so that’s amazing   right so we’ve got 3 000 plus people  have taken this what’s the current number  

24:16for patients who are testing probably say  syngap because i think earlier on you had   a slide around 28 but i feel like that  number’s gone up as well do you have that  

24:25at your fingertips sir that number has gone  up i don’t have it at my fingertips currently  

24:30uh i bel it’s i believe it was over a hundred  um uh that was less than two weeks ago but i’d  

24:36have to double check that but it’s gone up  significantly that’s awesome that’s awesome  

24:41yeah which is super exciting and the mozzie thing  like the ink is drying on the paper right so we   don’t even have a press release on that yet that’s  exactly right yeah the ink is drying as we speak  

24:52we should work on that press release oh yeah we  absolutely should i am so excited this is great  

24:59okay Corey that was my question  that you’re you’re your show   let’s see on the slide with the 20 different  diagnosis listed um to confirm out of the three  

25:13thousand submissions just have a hundred have  syngap diagnosis so we kind of answered that  

25:20just now um well the hundred was the the patient  population that was us that filled out um where  

25:28it says SYNGAP1 so that was actually our people  um in this list correct yeah so actually this is  

25:38uh the snapshot of all of those submissions so um  this would be the top diagnosis so these are the  

25:44patient diet patient reported diagnoses um so you  know when you fill out the symptom checker you um  

25:50can input any uh diagnosis that you’ve received  clinically um so of those so the question is  

25:55correct uh that’s right uh out of the 3 200 odd  uh submissions a hundred have self-identified  

26:02as having syngap um and then you can see the  prevalence of the other sort of top diagnoses   uh so to speak noted on the slide so that’s right  so we we found in addition to the hundred that we  

26:14already knew about we found the additional 100 um  patients that fit that profile out of the 3 200.

26:26all right and says next this is from ed in  the slide showing the symptoms of SYNGAP1  

26:34versus non-SYNGAP1 i assume that all other  disorders are lumped into the non-SYNGAP1 chart  

26:41if so have you looked into whether  they are any individual disorders   within that group that mirror SYNGAP1 more  closely than aggregate of all non-SYNGAP1

26:57so i guess i think i understand  this in temperatures one particular   phenotype that fits SYNGAP1 out of this bunch more  so than the rest of them yeah so that’s a really  

27:09interesting question um so the short answer is no  in that whether or not we’ve looked whether or not  

27:14we’ve run an analysis um that stratifies  non-SYNGAP1 submissions according to like  

27:23likeliness or likeness to the SYNGAP1 phenotypic  profile and the reason that we haven’t done that  

27:28is that a it would be quite uh intensive because  you’d have to sort of stratify uh you probably saw   that list of differential diagnoses and then do  a comparison which would involve retraining the  

27:38model whereas the approach here is to just sort of  use the knowns so what i mean by that is you know  

27:46the knowns for this training um for this program  and the way in which we set up the training  

27:52is that we had our one known be the syngap  and then had you know the sort of differential  

27:57diagnosis come in so if we wanted to run something  like what you’re proposing we’d have to make  

28:03sure that that initial step where we you know  did the really comprehensive set of symptoms   um for each of the fesian gap one and then the  comprehensive set for all of s differential  

28:12disorders um that we would do that set for every  uh diagnosis that we’d want to compare to sanger  

28:18which um you know obviously would be quite a few  um so the short answer to your question is no  

28:23uh we currently just have the sign got one  known comparison to the non SYNGAP1 submissions  

28:31um he said did any of them  previously have a different diagnosis  

28:37yes several um so i assume you mean the sign got  one um so that’s another really really rich um  

28:44you know source or resource both for us  and for the SRF so when patients can submit  

28:51to the symptom checker they can submit  uh you know obviously a really rich uh   amount of detail on their symptoms but they  also submit information on things like onset  

29:01um on things like severity and so what we can  map from that for the target condition is a very  

29:08very detailed um comprehensive almost we like  to call it the diagnostic odyssey right where  

29:14we can kind of say all right the majority of  one patients are experiencing an age of onset   around about this time uh the majority are getting  a diagnosis around about this time which you know  

29:24allows you to calculate kind of the the average  weight effectively for a diagnosis all of those  

29:29data points uh you know in many obvious ways very  useful um not only for the SRF but for everyone in  

29:35the signup community but uh certainly it’s because  it’s all patient initiated and patient reported  

29:41it’s a really powerful way of getting a snapshot  on what patients themselves um inpatient families  

29:46and caregivers are experiencing um and kind of  the ways in which they characterize their own  

29:52experience which i know the SRF is really excited  about and we found to be a really powerful tool  

29:57yeah i’ll tell you looking through the data  that we received it was very interesting   to see how many people had already had  some type of genetic testing in the past  

30:07but it has been you know 10 years five years  probably back when they weren’t testing for syngap  

30:15um so we’ve gone back and made a recommendation  to them to go back to the genetic company that did  

30:20the testing and get them to rerun the port is free  to them and there might be something even if it’s  

30:26not syngap something else might pop up that wasn’t  previously known at the time they had the um  

30:32original genetic testing done yeah that’s that’s  exactly it and that really underlines the power of  

30:38um you know the the kind of i guess sister slide  to this one where we’re um you know showing the  

30:43profiles of the patients that we’re saying hey  these you know didn’t submit a SYNGAP1 diagnosis   but they really you know the model’s telling  us they really maybe have a high likelihood of  

30:52um either being a patient or being very close to  or having you know a phenotypic profile that would   really benefit either from additional testing  maybe revisiting testing like you mentioned Corey  

31:01um or just you know additional resources and just  a little bit more to learn about um you know the   specifics of the disease so um in any of you know  regardless of which bucket you fall into in terms  

31:11of patient and diagnosis and what your previous  testing history was which we also captured by the   way um so yeah corey we were looking at that data  um in terms of you know the tests that previous  

31:21uh the patients have previously had no matter what  bucket you fall into we’re hoping that the program  

31:26is you know producing a resource that’s useful  for all of those patients yeah because so many  

31:31people don’t realize you can go back and have it  re-run they don’t realize the differences between   the different types of genetic testing um not  everyone tests for everything so it’s important  

31:42that they go back and really try to fight for that  diagnosis because so many people just don’t know  

31:47to ask the different questions that’s exactly it  yeah and it really speaks to you know there are  

31:53the srs um you know particularly good at  this but the role of patient advocacy orgs   of of that empowerment piece right so even if  we’re not saying oh you know you look a lot like  

32:03a sign get one patient that that the power of that  knowledge that um you know you can go through hey  

32:08you know you can ask your doctor to revisit this  you can ask your testing service to revisit this   or hey you know look at this gene or this variant  or whatever it might be those are the really kind  

32:16of empowering pieces um that this kind of  information can hopefully provide a link to

32:23i also wanted to make note we did create a  probably syngap facebook page so anybody that  

32:30got a high probability test result back from  the survey we made sure we’ve let them know  

32:37hey we’ve got a facebook page it’s  a probably syngap and you can kind  

32:42of go in here and and talk to other parents  that have also tested highly probable and uh  

32:49talk about your stories and different things  about your kids and it’s been really great to   for them to have a home too because they’ve been  so long without knowing and just it at the very  

32:58least finding people that have similar phenotypes  as their children has been so beneficial to them  

33:06yeah i mean that was just such an incredible um  piece of this of this program you know i know   that in in our you know we had weekly checkpoints  um as as we were running the program and certainly  

33:17hearing from the parents uh hearing the  testimonials that came in i mean we had   people who were using the program that had just  been you know like elbows deep in this diagnostic  

33:27odyssey for just years and years and years  um and like any light that you can shed there  

33:32um is just so impactful and powerful and so  being able to hear all of that through the SRF   and you know seeing the impact that you guys  are having and that we can have any small hand  

33:41in uh you know providing you with any tools or  resources that was definitely really powerful   for us um and i think just exactly what you said  Corey you know the the impact of information the  

33:51impact of community was definitely underlined by  by this program um and yeah i mean that growth  

33:57curve like mike said that growth curve speaks for  itself uh the srf community enormously engaged and  

34:02um yeah you guys are definitely a role model  for all the other all the other programs we’re   putting together we’ve been hanging out in the  undiagnosed parent groups a lot and you know it’s  

34:12funny you can kind of hang out in those groups and  see when they talk about their children um that  

34:19they seem to have very you know similar symptoms  and so we try to reach out and put the link in  

34:24there to let them know hey here’s a chance to to  maybe kind of narrow down that your odyssey so  

34:31um hopefully we’ll continue to find some more  people out of that sorry my dog went nuts  

34:37that’s okay i mean that’s actually a great segue  and i’m going to pounce at that opportunity corey   um to basically just make that exact point so you  know there’s a reason why i’ve just flipped this  

34:46slide again um obviously we’re incredibly excited  srf is really excited that we’ve signed a partner  

34:52um to connect up these two pieces you know the  patient finding and then the actual testing on   the diagnosis piece um which we’re really excited  about but you know soon that what that means in  

35:02terms of next steps and what your community is  going to see is that soon we’re going to have the   whole process connected up and you know your the  people within your community are going to start to  

35:11see the link pop up and they’re going to have that  opportunity to you know be considered for the free  

35:16genetic testing and you know we can only implore  you um and everyone in the community to do exactly  

35:22what you just mentioned right if there’s any  corner of you know the internet of social media of  

35:28any kind of you know disease influences and  whatnot um that you know you interact with during  

35:34your diagnostic odyssey that you can get get a  link to um or you know just share some information  

35:40with about the program that’s so powerful um  and that’s exactly what we need um because  

35:45you know that all we’re trying to do is all we can  do is develop mechanisms right and develop tools   and it’s really just uh the communities that  we serve so the communities like the srf  

35:55and patient communities that are the power behind  that so you guys really power the mechanism  

36:00um and it’s only you know without the  actual srf community we wouldn’t have   any of the results that i was just able to show  today and i think that you know the fact that  

36:08all of it all of our expectations are exceeded  um we’ve been able to produce this really you  

36:13know data rich really important sort of  powerful uh resource that’s giving us so  

36:19many insights on the on the community uh that’s  all down to you know the community itself it’s  

36:25you know we can develop all the mo tools  in the world but they’ve got to be used so   uh yeah we’re really excited to have this  opportunity come up and we’re excited to see  

36:31you guys put the link in every corner of the  internet once again one thing i would like to  

36:37you to speak on is the privacy um while we are  in these you know these undiagnosed parents  

36:44groups or epilepsy groups um there has been some  hesitancy for people to give their information you  

36:50know information is very critical nowadays  and their people are much more hesitant to  

36:56give out that data could you speak to a little  bit to ease some of the people that may be more   hesitant to fill out a survey because they’re  they’re worried about that information being sent  

37:06yeah absolutely um and and it’s a key point and  we really understand this we’ve got a couple of   um you know our team members who  have rare genetic conditions and so  

37:15have you know first-hand experience of um yes  you’re on that diagnostic odyssey and you’re   you’re really looking for answers but also you’re  you have that wariness right about your data so  

37:24um that data privacy has been centered in our in  how we’ve developed the symptom checker and also  

37:30all of our processes so from end to end um so what  that means is from the point at which a user even  

37:36clicks clicks um on a link to the symptom checker  or you know enters any information whatsoever  

37:41um we’re compliant with everything that you need  to be compliant with so the primary piece there   is hipaa um every single aspect of our system so  not only our you know website where you input any  

37:51information in the databases that we use those  are all hipaa compliant um which protects your   data and means that it’s all completely adherent  to um the relevant you know laws that protect any  

38:01personal um identifying health into information so  that is all very secure um in terms of where any  

38:07information is stored if at all um and that’s  end to end so that’s not only when you input  

38:12your information it’s also at any point you know  when your data is sent to our sequencing provider  

38:18to our report generate to our analysis it’s all  completely hipaa compliant all the way through   which isn’t it’s a bit of a gray area whether  that’s actually required but we’ve kind of gone  

38:26through like ironclad and made sure that every  single step in our process is compliant so there’s   that piece where you know on an institutional on  us on a legislative level we’re completely you  

38:36know almost extra compliant um but the second  piece is um you know our responsibility so we  

38:42take this really seriously um around how we  use data that we would have and there’s two key  

38:48pieces here the first is that anything uh that is  personally identifiable anything that could be in  

38:55any way linked back to you or anything like that  is completely um you know de-anonymized and also  

39:01is protected by consent so we will never share  any part of your data um without your explicit  

39:07consent um so nothing is assumed the consent has  to be explicit um and so that’s the first piece  

39:13around explicit consent of personal health data  and then the second piece is that any data that um  

39:19we do share so for example the data that i  just showed everyone um in this presentation  

39:24everything is anonymized um in a very uh stringent  way so for example any location information so  

39:31you know when you complete the symptom checker  that ip even is completely anonymized uh to zip  

39:37three level so uh you know the last two digits  of any zip is taken out so even those dots if  

39:42you zoomed in on that location map all you would  get is like a pretty sizable region um because   that would be the zip3 level so you wouldn’t be  able to see anything further than that so that  

39:51kind of thing as well we take really seriously um  so any kind of information that would be passed   on to any prospective partner even for example um  would be completely de-anonymized and again would  

40:00require that consent um so i think that those uh  i could probably talk about that for a long time  

40:06um but the two the two key pieces are  basically that we take it really seriously   um you know and that’s driven by the patients that  we have within our organization um and the second  

40:14piece is that we’re you know extra compliant um  with all of the regulations uh that that serve  

40:20uh health data that’s great thank you very much  for that all right and then we did have one more  

40:26question that came through um from an anonymous  attendee attendee she said um just to clarify how  

40:32many months were the 3 200 submissions over yeah  that’s a great point i can actually jump to that

40:41so it’s from i think about fab yeah there we go so   basically from uh early fab to june so what’s that  about four months yeah is that right yeah so it’s  

40:53it’s honestly it’s it’s astonishing growth um i  think by the end of the pilot uh like i mentioned   we had i think 600 so you can see the growth just  from that period alone uh we’re seeing submissions  

41:04continue to come in so uh you know we’re  still getting uh i think upwards of 50 a week  

41:10um and even that is like a slight drop off  you can see uh that the rate’s kind of uh   evened off a little bit in the in recent  days but that growth trend with that um  

41:19you know that the rate of growth is just is is  remarkable and definitely exceeded everything that   we thought would would happen so uh yeah about  four months to date great all right well i think  

41:31we are done with questions unless anybody has  the last minute when they want to pop in the chat

41:38we appreciate it abby and we  look forward to a question   finding out more and getting more data from  you guys that’s amazing it’s been great yeah  

41:48we’re so excited for next steps so um yeah  can’t wait for everything that’s about to   happen and uh yeah definitely echoing mike’s  excitement we’re really raring to go on this  

41:58well you have a wonderful day thanks  everybody you too thanks for having me