1
|
Omelchenko AA, Siwek JC, Chhibbar P, Arshad S, Nazarali I, Nazarali K, Rosengart A, Rahimikollu J, Tilstra J, Shlomchik MJ, Koes DR, Joglekar AV, Das J. Sliding Window INteraction Grammar (SWING): a generalized interaction language model for peptide and protein interactions. BIORXIV : THE PREPRINT SERVER FOR BIOLOGY 2024:2024.05.01.592062. [PMID: 38746274 PMCID: PMC11092674 DOI: 10.1101/2024.05.01.592062] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/16/2024]
Abstract
The explosion of sequence data has allowed the rapid growth of protein language models (pLMs). pLMs have now been employed in many frameworks including variant-effect and peptide-specificity prediction. Traditionally, for protein-protein or peptide-protein interactions (PPIs), corresponding sequences are either co-embedded followed by post-hoc integration or the sequences are concatenated prior to embedding. Interestingly, no method utilizes a language representation of the interaction itself. We developed an interaction LM (iLM), which uses a novel language to represent interactions between protein/peptide sequences. Sliding Window Interaction Grammar (SWING) leverages differences in amino acid properties to generate an interaction vocabulary. This vocabulary is the input into a LM followed by a supervised prediction step where the LM's representations are used as features. SWING was first applied to predicting peptide:MHC (pMHC) interactions. SWING was not only successful at generating Class I and Class II models that have comparable prediction to state-of-the-art approaches, but the unique Mixed Class model was also successful at jointly predicting both classes. Further, the SWING model trained only on Class I alleles was predictive for Class II, a complex prediction task not attempted by any existing approach. For de novo data, using only Class I or Class II data, SWING also accurately predicted Class II pMHC interactions in murine models of SLE (MRL/lpr model) and T1D (NOD model), that were validated experimentally. To further evaluate SWING's generalizability, we tested its ability to predict the disruption of specific protein-protein interactions by missense mutations. Although modern methods like AlphaMissense and ESM1b can predict interfaces and variant effects/pathogenicity per mutation, they are unable to predict interaction-specific disruptions. SWING was successful at accurately predicting the impact of both Mendelian mutations and population variants on PPIs. This is the first generalizable approach that can accurately predict interaction-specific disruptions by missense mutations with only sequence information. Overall, SWING is a first-in-class generalizable zero-shot iLM that learns the language of PPIs.
Collapse
Affiliation(s)
- Alisa A. Omelchenko
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Computational and Systems Biology, School of Medicine, University of Pittsburgh, PA, USA
- The joint CMU-Pitt PhD program in computational biology, School of Medicine, University of Pittsburgh, PA, USA
| | - Jane C. Siwek
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Computational and Systems Biology, School of Medicine, University of Pittsburgh, PA, USA
- The joint CMU-Pitt PhD program in computational biology, School of Medicine, University of Pittsburgh, PA, USA
| | - Prabal Chhibbar
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Integrative systems biology PhD program, School of Medicine, University of Pittsburgh, PA, USA
| | - Sanya Arshad
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
| | - Iliyan Nazarali
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
| | - Kiran Nazarali
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
| | - AnnaElaine Rosengart
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
| | - Javad Rahimikollu
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Computational and Systems Biology, School of Medicine, University of Pittsburgh, PA, USA
- The joint CMU-Pitt PhD program in computational biology, School of Medicine, University of Pittsburgh, PA, USA
| | - Jeremy Tilstra
- Department of Immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Division of Rheumatology and Clinical Immunology, Department of Medicine, School of Medicine, University of Pittsburgh, PA, USA
| | - Mark J. Shlomchik
- Department of Immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
| | - David R. Koes
- Department of Computational and Systems Biology, School of Medicine, University of Pittsburgh, PA, USA
| | - Alok V. Joglekar
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Computational and Systems Biology, School of Medicine, University of Pittsburgh, PA, USA
| | - Jishnu Das
- Center for Systems immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Immunology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, USA
- Department of Computational and Systems Biology, School of Medicine, University of Pittsburgh, PA, USA
| |
Collapse
|
2
|
Kim KH, Pereira NL. Genetics of Cardiomyopathy: Clinical and Mechanistic Implications for Heart Failure. Korean Circ J 2021; 51:797-836. [PMID: 34327881 PMCID: PMC8484993 DOI: 10.4070/kcj.2021.0154] [Citation(s) in RCA: 11] [Impact Index Per Article: 3.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/08/2021] [Accepted: 06/14/2021] [Indexed: 11/11/2022] Open
Abstract
Genetic cardiomyopathies are an important cause of sudden cardiac death across all age groups. Genetic testing in heart failure clinics is useful for family screening and providing individual prognostic insight. Obtaining a family history of at least three generations, including the creation of a pedigree, is recommended for all patients with primary cardiomyopathy. Additionally, when appropriate, consultation with a genetic counsellor can aid in the success of a genetic evaluation. Clinical screening should be performed on all first-degree relatives of patients with genetic cardiomyopathy. Genetics has played an important role in the understanding of different cardiomyopathies, and the field of heart failure (HF) genetics is progressing rapidly. Much research has also focused on distinguishing markers of risk in patients with cardiomyopathy using genetic testing. While these efforts currently remain incomplete, new genomic technologies and analytical strategies provide promising opportunities to further explore the genetic architecture of cardiomyopathies, afford insight into the early manifestations of cardiomyopathy, and help define the molecular pathophysiological basis for cardiac remodeling. Cardiovascular physicians should be fully aware of the utility and potential pitfalls of incorporating genetic test results into pre-emptive treatment strategies for patients in the preliminary stages of HF. Future work will need to be directed towards elucidating the biological mechanisms of both rare and common gene variants and environmental determinants of plasticity in the genotype-phenotype relationship. This future research should aim to further our ability to identify, diagnose, and treat disorders that cause HF and sudden cardiac death in young patients, as well as prioritize improving our ability to stratify the risk for these patients prior to the onset of the more severe consequences of their disease.
Collapse
Affiliation(s)
- Kyung Hee Kim
- Division of Cardiology, Incheon Sejong General Hospital, Incheon, Korea.
| | - Naveen L Pereira
- Department of Cardiovascular Medicine, Mayo Clinic, Rochester, MN, USA.,Department of Clinical Genomics, Mayo Clinic, Rochester, MN, USA
| |
Collapse
|
4
|
Meghji Z, Nguyen A, Miranda WR, Geske JB, Schaff HV, Peck DS, Newman DB. Surgical septal myectomy for relief of dynamic obstruction in Anderson-Fabry Disease. Int J Cardiol 2019; 292:91-94. [PMID: 31262606 DOI: 10.1016/j.ijcard.2019.06.042] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/16/2019] [Revised: 05/23/2019] [Accepted: 06/18/2019] [Indexed: 11/28/2022]
Abstract
Patients with Anderson-Fabry Disease (AFD) and severe left ventricular hypertrophy complicated by left ventricular outflow tract (LVOT) obstruction may benefit from surgical septal myectomy (SSM). Mid- and late outcomes following surgery have not been established, and we sought to better characterize postoperative outcomes following septal myectomy. Between January 2011 and June 2017, 7 patients (6 females) with AFD underwent SSM. The median (range) age at the time of surgery was 53 (37-72) years; 4 patients had a positive family history of AFD and a preoperative diagnosis of AFD. Extracardiac features suggestive of AFD were present in 3 patients and all but 1 (female) had reduced α-galactosidase A activity. All patients had severe left ventricular hypertrophy and LVOT obstruction on transthoracic echocardiography. Preoperatively, all patients were severely symptomatic with New York Heart Association (NYHA) class III symptoms. There was no early mortality following surgery. The median in-hospital length of stay was 5 (4-7) days with 6 patients reporting NYHA class II or less symptoms at 3 month follow-up. Long-term outcomes were favorable with 4 patients reporting sustained NYHA class II or less symptoms, but 2 patients had recurrence of NYHA class III symptoms without evidence of recurrent LVOT obstruction. In conclusion, SSM appears to provide favorable short- and long-term relief of severe, symptomatic LVOT obstruction but may not alter progression of Fabry cardiomyopathy.
Collapse
Affiliation(s)
- Zahara Meghji
- Department of Cardiovascular Surgery, Mayo Clinic, Rochester, MN, United States of America
| | - Anita Nguyen
- Department of Cardiovascular Surgery, Mayo Clinic, Rochester, MN, United States of America
| | - William R Miranda
- Department of Cardiovascular Diseases, Mayo Clinic, Rochester, MN, United States of America
| | - Jeffrey B Geske
- Department of Cardiovascular Diseases, Mayo Clinic, Rochester, MN, United States of America
| | - Hartzell V Schaff
- Department of Cardiovascular Surgery, Mayo Clinic, Rochester, MN, United States of America
| | - Dawn S Peck
- Department of Biomedical Genetics, Mayo Clinic, Rochester, MN, United States of America
| | - Darrell B Newman
- Department of Cardiovascular Diseases, Mayo Clinic, Rochester, MN, United States of America.
| |
Collapse
|