Prediction of AL Amyloidosis Using Machine Learning

AL amyloidosis is a rare protein disorder that can be often fatal if it is not diagnosed and managed early. This disorder is caused by misfolding of proteins that clump together and form amyloid fibril deposits in major body organs. Diagnosis of AL amyloidosis is often not easy as the signs and symptoms can be mistaken for common diseases. The median survival rate after diagnosis is less than six months when the underlying plasma cell dyscrasia is left untreated in AL amyloidosis patients. Hence, it is paramount to develop novel diagnostic methodologies that are not based on the signs and symptoms of AL amyloidosis, but are based on the underlying molecular mechanisms of amyloidogenic clone, which can provide evidences for predisposition much long before the disease sets its course on the body. We propose to develop a machine learning technique to predict whether a light chain amino acid sequence will form proteins that misfold and produce amyloid fibrils, leading to AL amyloidosis.

Faculty Supervisor:

Bin Ma;Lila Kari

Student:

Anupa Murali

Partner:

Rapid Novor Inc

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

University of Waterloo