Good afternoon. I hope you all can hear me now. To begin with, I would like to tell you about the ACMG Annual Meeting that is happening in Charlotte, North Carolina from April 10 to 14. So we would like to invite you to join us and register today in that link that you can see here. And we would like to welcome you to the ACMG Genomics Case Conference from today, hosted by Johns Hopkins University and the Greater Baltimore Medical Center. We are going to talk about much-needed exchange, solving unsolved cases. The presenters today are me, Nara Sobreda from Hopkins, and Tony Kline from the Greater Baltimore Medical Center. And I don't have any disclosures. Tony's disclosure is that he has a salary employment with the Cornelia de Lange Syndrome Foundation. There is no disclosures, or these are the disclosures from the ACMG team. This case conference is intended for learning purpose only. It is one CME credit. And if you register, you will get a certificate from your CME. These are the overall learning objectives of the case conference today. And these are the learning objectives for this session for today. You can send your questions and answers to Anne throughout the talk. And we will answer verbally at the end of the presentation. So now I'm going to pass to Tony, and she will start the presentation from today. Hello, everybody. Today we are talking about solving unsolved cases. So first, I'm going to describe to you my patient from Baltimore, whom I followed since birth. And at birth, he had polydactyly of the hands, foot deformity, right cryptorchidism, extra nipples, and a prominent metopic suture that you can see in most of these pictures. In a workup after birth, he was found to have aortic dilatation and pulmonic stenosis, as well as hydronephrosis, and then was found to have a neurogenic bladder and vesiculoureteral reflux. On exam, he was found to have optic nerve atrophy. And over time, it was clear that he developed intellectual disability. And here's some pictures. You can see the polydactyly and the overlapping toes. And in addition, we noted a vertical groove in his tongue with a bifid tip, as you can see here. He also was born with a hip dysplasia and developed scoliosis. And on MRI, as you can see on the right, he had some segmentation defects. He has joint hyperextensibility and hypotonia and hyporeflexia. He has developed a moderate range of intellectual disability, and he is nonverbal. He has ADHD. And more recently, he has developed seizures. So I followed him since birth. We did a workup over time, and he was not found to have anything specific. He was seen at the Undiagnosed Disease Clinic at the NIH, and they did not diagnose any specific condition. We wondered if he could have Kabuki syndrome, and he was seen at Hopkins in the Kabuki Clinic, and he had negative results. He was seen at the Kabuki Clinic, and he had negative testing. Throughout his early childhood and mid-childhood, his mother would frequently ask me about him appearing on mystery diagnosis on the Discovery Channel. I presented him at the David Smith Workshop, and people felt that he did not likely have Kabuki syndrome, and they recommended whole exome sequencing, which his insurance denied. So we enrolled him in the Baylor Hopkins Center Genomics Project in order for him to get this test. And this shows you a list of some of his workup. He did have hypomyelination and a small posterior fossa on MRI, but the rest of the studies were negative, including gene sequencing for the genes shown here. And now Nada is going to continue. Nara, we can't hear you. Sorry. I forgot to unmute. Can you hear me now? Yes. Hello? So the Baylor Hopkins Center for Mendelian Genomics is an NIH-funded project that sequenced the whole exome sequencing of patients with rare Mendelian phenotypes, and this is what we did with the patients that Tony submitted to our project. And if you don't know a lot about the Baylor Hopkins Center for Mendelian Genomics, you can go to that address and learn a little bit more about it, and if you are interested, contact us to submit your patients to the project. So when we got the case from Tony, we decided to sequence the trio, and we did the analysis looking for the novel variants and looking for the recessive homozygous or compound heterozygous variants. And this is actually one of the few variants that appeared as the final candidate gene, and the variant in this gene was a missense on the last code of exome 5 that predicted to abolish the spice donor site. So we confirmed by Sanger sequence the variant in the patient, and we confirmed that the parent didn't have the variant. And we did a little bit more of research on the gene. We found that there was no loss of function variants in this gene described in EXAC or in the exome variant server. There was no deletions reported in this cipher involving this gene. So these are among the few candidate genes that we had at the end of our analysis, our best candidate genes. And we talked to Tony because at this point, we wanted to find other patients with similar phenotypes and variants in the same gene. And the phenotype that you just heard was a very unique phenotype, and we could not really find patients with this similar phenotype anywhere that we had access to. So we thought that this would be a perfect case for gene matches. So I don't know if you all know about GeneMatcher, but GeneMatcher is a database that we developed, the Baylor Hopkins Center for Mendelian Genomics developed, exactly trying to solve our unsolved case, the case where we had a few candidate genes, and we needed another patient with a similar phenotype and mutation in one of these genes to select the main gene or the causative gene among these few ones in the list. So in GeneMatcher, you can, most of our users use GeneMatcher to match on a gene name, but you also can match on an on-mean phenotype ID if you have a candidate diagnosis for your patient. You can match on genomic location and on features. When you create an account in GeneMatcher, you can, you will tell us if you're a researcher, a healthcare provider, or a patient, and then you can actually select if you want to restrict your matches, you can select who you want to match to. So you can say that I don't want to match a patient, for example, if you don't feel comfortable exchanging information directly with patients. As of March 1st, there were 8,646 genes in GeneMatcher from 4,871 submitters from 71 counties. You can see here on this graph how the number of genes and the number of matches and the number of genes that matches have been growing in the database. We have had 50,050 matches on 4,048 genes. So you can see that less than half of the genes in GeneMatcher have had a match. We have a few more than 4,000 genes associated to a phenotype by one, probably, mostly by the whole exome sequencing, and that have never matched to anybody else in the database. So we have a lot of work to do in terms of trying to find the phenotype associated to these genes. In this table, you can see the number of genes and how many times they have matched in our database. So we have 4,204 genes that never matched, 1,876 that matched one time, down to two genes that matched 29 times and two genes that matched 33 times. So you can see that the number of genes that match a lot is small, but it's also helpful because if you have one of these two genes that matches 33 times, you probably know by now that this is not a good candidate for your patient. The same thing, you have your one gene that matches 59 times or 63 times, you know that this is probably not one of the best candidates in your candidate list. We are doing more work on this table, in these numbers, to try to understand better what's the behavior of our matches. And as we get this data out, we get this data, we put it out. So you can go to GeneMatcher in our statistics page and get many of these numbers that I'm telling you about. We have had, as of today, 88 publications that cite GeneMatcher or GeneMatcher match. So we don't have a way to monitor our publications and our matches because every data that is in GeneMatcher is private to the user. So I don't know what are the genes that are in GeneMatcher. I only know what are the genes that I put in GeneMatcher. So I cannot easily follow the publications. The only way that I have to follow the publications is if the users actually cite GeneMatcher in the paper when they are in the writing process. And this is how we put the papers in here. So GeneMatcher is part of the Matchmaker Exchange. And I'm going to tell you a little bit of the Matchmaker Exchange starting the next slide. But from here, I want to show you that you can actually select the database that you want to send your query to. So you have the Australian Genomics Health Alliance, the Broad Matchbox, Decipher, MyGene2, and Phenome Central. You can have your data in GeneMatcher and from there, send your data to any of these as you choose. And you can also choose on the right side, you can see what data you want to send to each one of these databases. Only the gene name, the gene name and the genomic location, a gene name and some features of your patient, you choose it. And you have the option also above this table here you can see, save this Matchmaker Exchange match and rerun it every month. So you don't have actually to go back every so often and redo the query. You can actually check this box and the query will be done for you automatically every month in these other databases. So as I was telling you, the Matchmaker Exchange, it's a group of clinicians and researchers that got together and we were each working on different database with phenotype and genomic information on the patients that we were working on and working with. And we decided that it would be a good idea to get this database to talk to each other in a way that I as a user didn't need to go to each one of these databases to query them. So I could go to one, open an account in one and from there, query the other database. So we now have seven databases connected through our API. And you can see here what the databases are, Phenome Central, MyGen2, Patient Archive is the Australian one, Monarch, it's an animal model database, Matchbox and Decipher. So from there, from each one of them, you can actually send your data to query the other database. Each one of these databases stores and shares a slightly different kind of data, phenotype data and genotype data, but in a slightly different way. So if you go to the website that you can see here on the top, the matchmakerexchange.org, you can have a little bit more of details on what kind of data you can store in each one of these databases and what kind of data you can share from this database to the others and vice versa. So this is very important because depending on the kind of genomic data and phenotype data that you have and that you can share, one or the database may suit you better. So it's important if you don't have an account in one of them already and you want to create now or in the future, it's important to check it out in these websites. We are always updating the information there as the website and the database change. And it has changed. We each started in a slightly different way and kept adding functionalities to each one of the databases. So it's always important to go there to this matchmakerexchange.org website and check what's going on with each one of the databases at the time when you're interested in. So I told you a little bit about GeneMaster. I'm going to tell you a little bit about Phenome Central. So in Phenome Central, you also can store phenotype data, genotype data, and you can actually upload your VCF file there. And they will use a pipeline to come up with scanned genes for your VCF file. And then you will choose if you want to share these scanned genes or not with other databases or try to find other match inside the Phenome Central also. So Phenome Central has, as of now, 1,100 users, 4,200 cases with 1,400 unique genes, 800 exome sequences. They have 50,000 HPO annotations. And they have three published papers and eight papers in preparation with matches involving Phenome Central. And Phenome Central is based in Canada. Then we have the Matchbox. Matchbox is based from the Broad. And they are not open access in a way that you cannot go there and create your account and put your data in there. But you can go through all the databases and query Matchbox. So if you have the data in Phenome Central, you can send your data from Phenome Central to Matchbox and query the data that is in there. And the data that is in there, data from projects that are being sequenced by the Broad, and they have 17,120 cases as of now. They have shared the data through the matchmaking exchange of 182 cases. These are corresponding to 250 genes and 481 HPO terms. So they have had, by using MME, 19,180 matches. And they have two publications from this matching that I think, if they are not out already, are just coming out now. Decipher, I think most of you know about Decipher very well. Decipher is the oldest of these databases. It's based in the UK, used in more than 40 countries. It has more than 25,000 open access patient records, more than 75,000 phenotypes, more than 4,000 sequence variants, more than 29,000 copy number variations. And they share semi-anonymized patient data since 2014. And they have more than 1,500 publications. They have been facilitating identification and interpretation of phenotype-linked pathogenic genetic variation in red dissolvers. I'm going to focus in this slide. I know that this is small, but I want you to focus on the lower half where they show where you are trying to find another match to your patient inside the database. And you can also go from the database to the other matchmaking exchange databases. So you can see here Phenome Central, GeneMatcher, MyGene2, and the Matchbox. And you can query them with your data and see if you get a match. MyGene2 is together with GeneMatcher. These are the only two databases as of now that allow patient participation. But MyGene2 is much more focused on patient participation than GeneMatcher. We have much less patient data in GeneMatcher than they have in MyGene2. They also gather data from clinicians and researchers. They also gather phenotype information and genomic information. You can upload your VCF files in there or your BAM files. And they will use Mutalizer to come up with a candid list of genes. And then you can decide how to share these genes inside the database and with other databases connected through the MME. So if you put a description of your patient, the database will get and select the terms describing this patient that they think are more appropriate to make the phenotype profile of your patient. And this is a candid list of the genes from one of their cases. And then you can select these genes to match inside the database and with other databases through the MME. And here you have a summary of the data in MyGene2 as of now, 1,323 profiles for more than 880 user accounts. About 30% of them are submitted by families and 70% submitted by clinicians or researchers. And then you have 807 unique genes, approximately 50% of them submitted by families. And you can see here on the right side the novel unpublished candid genes shared by families. And then you can have it posted by family to MyGene2 before publication. So this is a very good resource for the patients. And every time I get a query from a patient, I always send the patient towards MyGene2 because I think it's very good for them also. Then you have the Australian Genomics Health Alliance where you also can store and share phenotype data, the diagnosis ID, and the gene name. In this case here, the user only had phenotype information. They didn't have diagnosis of genes. But they shared this data to the MME. And they got, you can see, two matches, one with Phenome Central that matches 50% of the phenotype and one with GeneMatcher that matches 100% of the phenotype. One thing that I wanted to bring the attention to is that some of the databases do not allow match on phenotype only. Most of the databases require phenotype and a gene. You can match on a gene only, but phenotype only is still really complicated to ask to match because it can generate a lot of false positives and be very unspecific. So if all you have about your patient is that they have intellectual disability, you can imagine how many false positives matches you are going to get or unspecific matches you are going to get. So now I'm going to pass it back to Toni, and she will tell what happened to the head case once we put it in GeneMatcher. So after Nada put it into GeneMatcher, a year and a half went by. And then I got an email from my colleague, Dr. Mike Ennis, in Calgary in Canada. And he said that he had just put a change in a gene into GeneMatcher and instantly gotten a match that I had a similar patient. So his patient was 13 at the time, and this is how he looked when he was younger and a little bit older. He also has intellectual disability. He was born with a craniosynostosis that needed surgery. He developed a marfanoid habitus with a long face that I'll show you in a minute. He has proptosis and ptosis, down-planting palpebral fissures, a prominent nasal bridge, hypoplastic alias nasi, and a long philtrum. And this is how he looked when he was a little older. He also was found to have a vertical groove in the tongue, and he has malocclusion and an open mouth habitus as well as a quite high palate and a bifid uvula. He also has low set ears with underdeveloped helices and scoliosis and hip dysplasia. And his scoliosis is also due to segmentation defects, as you can see on the MRI. And this is a picture of his hands and feet that you may remember look a little similar to my patient. They did a workup of many organ systems. He has a normal MRI of the brain, and in addition to the segmentation defects, has a syrinx and a terminal lipomyelomeningocele. He was born with a VSD that – actually two VSDs that closed spontaneously. And they also had done an extensive workup, had found CNV on an array, and did sequencing of similar genes that were negative. When we compared the pictures of our two patients, these were the common facial features. They both had ridging of the metopic suture, long down-slanting palpebral fissures, long faces as they got older, underdeveloped ear helices with a wide nasal bridge and hypoplastic alias nasi, an open mouth habitus with the corners turned down. They both had a high palate, and they both had a prominent midline groove of the tongue with a bifid tip. And when we further discussed their clinical features, both patients had a very high pain tolerance, decreased sweating, so we thought dysautonomia symptoms, intermittent rash, and poor motility of the GI system with constipation. So we put our patients into an article along with Nada and her colleagues, and this was written with the first author, Billy Oh, who's also in Calgary. And when this was put into the OMIM catalog, they called it Oh-Klein syndrome. And since this paper was published, additional patients have been brought to our notice. So in this slide, the first two patients, cases three and four, were previously published before our article, and these were deletions on the long arm of chromosome 9 encompassing the HnRnPK gene. And you can see some similarities with their faces. And so they both had a microdeletion. I neglected to say that the Calgary patient was found to have a frameshift mutation in the HnRnPK gene. The next patient was found to have a splice site variant, and case six was subsequently published by a group in the U.K., and they also have a frameshift variant. And you can see this patient when he was younger. You can see some excessive skin in the back of his neck. And then here on the left, you can see as he's getting older, he develops a longer face. Case seven I also saw in my clinic in Baltimore, and this patient has a nonsense variant. And the last baby was found to have a frameshift variant. So overall, the facial features in common include prominent metopic suture, redundant skin around the eyes and neck in some patients, longer faces with aging, hypoplastic alinase, as you can see in these two noses, an open mouth habitus from early childhood, an M-shaped cupid's bow on the upper lip, as you can see in the top two pictures, and the vertical groove in the tongue, and some patients had the bifid tip. Other findings in common include craniosynostosis, cardiac defects of all types, genitourinary tract defects that end up causing the greatest morbidity in these patients, vertebral segmentation anomalies that can lead to scoliosis. All patients have had hypotonia, and all patients have had moderate to severe intellectual disability. We have been able to make specific clinical recommendations when found to have a new patient with this, and these include an echocardiogram, a renal ultrasound and VCUG if clinically indicated, x-rays of the spine and the hips, MRI of the brain and the spine, an eye exam, hearing assessment throughout childhood, and monitoring development with referral to early intervention therapy. In addition, monitoring for craniosynostosis when the patients are infants. Once the two families heard about each other and several more were found, they formed a Facebook group which is called the O'Klein Syndrome Support Group, and there's currently 31 members. That includes some parents and family members, and the patients have been able to exchange clinical information and family or social information. And through this group we have planned our first international O'Klein Syndrome family gathering that will be in Baltimore on July 21st this summer, and families are searching for ways to help fund travel to be able to come to this gathering. And now I'm going to turn it back over to Nada. We have here a list of references that we put together, and I think Claudia put it together as a handout, so you can go through this reference and learn more about Matchmaker Exchange, GeneMetrics, the other database connected to it. I think someone needs to mute because I'm getting a lot of echo. Hello? Okay. So you can read more about Matchmaker Exchange and GeneMetrics and the other database connected through MME. You can also learn more about the O'Klein Syndrome, and you can always contact us if you want to learn more about the syndrome or the Matchmaker Exchange and each one of these databases connected through MME. I would like to thank you all for your attention and all the people who work with us and make this work possible, in the Baylor Hopkins Center for Mendelian Genomics, in the Matchmaker Exchange team, and Tony for participating and everybody else involved in the publication of the patient. That actually made it possible for us to get information from many other patients, and the Facebook page is actually, as of now, I think, the better source of information about the phenotype of this novel syndrome that we have in the moment. So this is all the power of sharing your data, sharing your phenotype data, your genomic data, and the patients also sharing that data. They also want to share. They go to Facebook, and they share the information and their struggles, and you can see through the MyGen2 database that they are really interested in sharing the data and getting answers for their problems. So I'm now going to answer the questions that I have here. The first question is, did you shortlist Alkline syndrome as you were going for Matchmaker? So the Alkline syndrome was described because of the match. Before the match, the phenotype that Tony had put into the Baylor Hopkins, nobody knew what it was, and nobody knew what gene was associated to it, and the gene, when we found it in our exome sequencing analysis, had not been associated to any disease before also. So all this information was able to come out because of the match that happened through Tony's patient and the patient from Canada through GeneMatcher. And then with the publication, they found more patients, and now the Facebook page is finding more patients. So there was no Alkline syndrome at the time when we did the analysis and at the time when you put the data in GeneMatcher. I hope this answers your question. Then the next question is, did patients pass newborn hearing screening? Why was hearing assessment recommended? So, Tony, I'm going to pass that one for you. Okay, so the patients did pass the newborn hearing screening, but 40% of the patients that we have seen to date have had hearing loss. Half of them have had conductive hearing loss, and half have had sensorineural hearing loss. And I neglected to say that we're doing revisions on a paper with additional patients and a review of the phenotype that should be coming out soon. And then the next question that I have here is, what proportion of unsolved cases are solved by dominant de novo variants in genes no previous linkage to disease? So this is actually a very good question. I don't think we know the answer. I certainly can get that number. If I go back to the 88 publications from GeneMatcher, I can figure out how many of them are de novo mutations and how many of them are homozygous and how many of them are compound heterozygous. This is actually a good exercise for us to do and a good information for us to get from our publication list. I'm going to do that. But as of now, I don't know the answer for that. I don't see any other questions here. Does anybody have any additional questions? Since we have a little bit of time, I can tell you also something else about the Matchmaker Exchange is that sometimes in the databases, and not all of them but in GeneMatcher for sure, you can get connected to researchers who are actually working animal models for the gene that you are interested on. And it has been a source of positive matches and publications describing evidence for calling the gene causative because you found an animal model for the gene that you are interested on. So we do have some researchers that have fly models and mouse models that use GeneMatcher. So this is another thing that we have there. So if no other questions, I think I want to thank you all for being here today, and I'm glad to answer any questions by email later whenever you are interested on participating in one of the databases linked through the MME.

ACMG Genomics Case Conference

Johns Hopkins University

Greater Baltimore Medical Center

unsolved cases in genomics

Nara Sobreda

Tony Kline

Oh-Klein Syndrome