US 12,393,187 B2
Systems and methods for identifying subcomponent failure in sequencing instruments
Yasmin Sahaf, San Diego, CA (US)
Assigned to ILLUMINA, INC., San Diego, CA (US)
Filed by ILLUMINA, INC., San Diego, CA (US)
Filed on Mar. 31, 2022, as Appl. No. 17/710,665.
Application 17/710,665 is a continuation of application No. PCT/US2022/021813, filed on Mar. 24, 2022.
Claims priority of provisional application 63/169,676, filed on Apr. 1, 2021.
Prior Publication US 2022/0317674 A1, Oct. 6, 2022
Int. Cl. G06N 5/01 (2023.01); B01L 3/00 (2006.01); G05B 23/02 (2006.01)
CPC G05B 23/0281 (2013.01) [B01L 3/502715 (2013.01); G05B 23/0229 (2013.01); G05B 23/0272 (2013.01); G06N 5/01 (2023.01)] 36 Claims
OG exemplary drawing
 
1. A computer-implemented method of diagnosing system malfunction and isolating a cause of the system malfunction among a plurality of replaceable subcomponents, the method including:
applying, to time series data from at least one sequencing run, preprocessors that detect time series abnormalities, wherein time series abnormalities include discontinuities, drift, lack of expected correlation, and trends, wherein the discontinuities are detected in focus scores by:
applying a smoothing filter to the focus scores time series per image channel in the sequencing run;
calculating magnitudes of differences between the smoothed focus scores at successive cycles in the focus scores time series per image channel in the sequencing run;
detecting a number of successive cycles where the magnitudes of differences between successive cycles exceeds a predefined threshold;
for at least one image channel in one sequencing run, feeding at least part of the output of the preprocessors to a trained tree-based classifier and receiving a classification of the at least one sequencing run as abnormal;
feeding at least part of the output of the preprocessors for the abnormal sequencing run to an expert rule system and receiving a root cause isolation of the system malfunction to a particular subcomponent; and
generating at least one notification of the particular subcomponent causing the system malfunction for display on a graphical user interface, wherein the at least one notification comprises, based on the detecting the number of successive cycles where the magnitudes of differences between the successive cycles exceeds the predefined threshold, an indication of a malfunction of a z-stage subcomponent used to correctly position an imager relative to a flow cell.