A molecular-based approach using long, non-coding RNA and enhancer-associated lncRNA gene expression signatures to classify multiple sclerosis using peripheral whole blood
ECTRIMS Online Library. Spurlock, III C. 10/26/17; 199954; P299
Charles F. Spurlock, III
Charles F. Spurlock, III
Contributions
Abstract

Abstract: P299

Type: Poster

Abstract Category: Clinical aspects of MS - 1 Diagnosis and differential diagnosis

Background: The vast majority of the human genome is transcribed, but less than 3% of genomic space is occupied by known protein-coding gene exons. New RNA classes have been discovered including long, non-coding RNAs (lncRNAs) and enhancer-associated lncRNAs (eRNAs) that play pivotal roles in the development of the immune system and regulation of innate and adaptive immune responses. Recently, we found these RNAs are differentially expressed in autoimmune disease and are in linkage with genetic variants that confer disease risk. lncRNA and eRNA expression differences were observed in different stages of MS. Measuring these changes compared to control groups could provide useful diagnostic information for healthcare providers.
Objective: To compare machine learning classifiers using messenger RNA (mRNA), lncRNA, and eRNA expression data derived from whole blood to classify multiple sclerosis (MS) and distinguish MS from other inflammatory and non-inflammatory neurologic diseases.
Methods: Whole blood collected into PAXgene tubes was obtained from healthy subjects (n=446), clinically isolated syndrome patients who transitioned to clinically definite MS (n=84), MS patients prior to treatment (n=90), established MS patients (n=412), and other inflammatory (n=132) and non-inflammatory (n=115) neurologic diseases. RNA sequencing was performed using a subset of healthy control and MS patient samples to derive three sets of 48 highly differentially expressed gene candidates across mRNA, lncRNA, and eRNA gene targets. RT-PCR was performed on all patients recruited (n=1,279) and a portion of the datasets generated were used to train and independently validate machine learning classifiers capable of identifying MS.
Results: We found that differences in expression of mRNAs and known lncRNAs ranged in magnitude from two-fold to eight-fold across our study cohorts while differences in expression of eRNAs ranged between four-fold to 1,000-fold. Compared to mRNAs, greater expression differences observed in lncRNAs and eRNAs increased confidence and discriminatory power of machine learning predictions with sensitivity and specificity exceeding 90% across case/control comparisons.
Conclusions: lncRNA and eRNA expression data from whole blood analyzed using machine learning could provide clinically actionable, diagnostic information to healthcare providers.
Disclosure: C.F. Spurlock, III and T.M. Aune are co-founders and shareholders of IQuity, Inc. J.T. Tossberg has a financial interest in IQuity, Inc. All other authors declare no significant financial interest. This work was funded by IQuity, Inc. and the National Institutes of Health (USA).

Abstract: P299

Type: Poster

Abstract Category: Clinical aspects of MS - 1 Diagnosis and differential diagnosis

Background: The vast majority of the human genome is transcribed, but less than 3% of genomic space is occupied by known protein-coding gene exons. New RNA classes have been discovered including long, non-coding RNAs (lncRNAs) and enhancer-associated lncRNAs (eRNAs) that play pivotal roles in the development of the immune system and regulation of innate and adaptive immune responses. Recently, we found these RNAs are differentially expressed in autoimmune disease and are in linkage with genetic variants that confer disease risk. lncRNA and eRNA expression differences were observed in different stages of MS. Measuring these changes compared to control groups could provide useful diagnostic information for healthcare providers.
Objective: To compare machine learning classifiers using messenger RNA (mRNA), lncRNA, and eRNA expression data derived from whole blood to classify multiple sclerosis (MS) and distinguish MS from other inflammatory and non-inflammatory neurologic diseases.
Methods: Whole blood collected into PAXgene tubes was obtained from healthy subjects (n=446), clinically isolated syndrome patients who transitioned to clinically definite MS (n=84), MS patients prior to treatment (n=90), established MS patients (n=412), and other inflammatory (n=132) and non-inflammatory (n=115) neurologic diseases. RNA sequencing was performed using a subset of healthy control and MS patient samples to derive three sets of 48 highly differentially expressed gene candidates across mRNA, lncRNA, and eRNA gene targets. RT-PCR was performed on all patients recruited (n=1,279) and a portion of the datasets generated were used to train and independently validate machine learning classifiers capable of identifying MS.
Results: We found that differences in expression of mRNAs and known lncRNAs ranged in magnitude from two-fold to eight-fold across our study cohorts while differences in expression of eRNAs ranged between four-fold to 1,000-fold. Compared to mRNAs, greater expression differences observed in lncRNAs and eRNAs increased confidence and discriminatory power of machine learning predictions with sensitivity and specificity exceeding 90% across case/control comparisons.
Conclusions: lncRNA and eRNA expression data from whole blood analyzed using machine learning could provide clinically actionable, diagnostic information to healthcare providers.
Disclosure: C.F. Spurlock, III and T.M. Aune are co-founders and shareholders of IQuity, Inc. J.T. Tossberg has a financial interest in IQuity, Inc. All other authors declare no significant financial interest. This work was funded by IQuity, Inc. and the National Institutes of Health (USA).

By clicking “Accept Terms & all Cookies” or by continuing to browse, you agree to the storing of third-party cookies on your device to enhance your user experience and agree to the user terms and conditions of this learning management system (LMS).

Cookie Settings
Accept Terms & all Cookies