The overall scanning signal of the expression profiling microarray is relatively low. Is this due to batch-to-batch variation of the microarray or an operational issue?

If the overall signal is half as low and the internal control is also low, the first priority is to check the scanner PMT and laser, as this is a systemic issue. First, rescan the same area using a previously scanned chip from the last batch that had normal signal to see whether the signal values recover. If they return to normal, it indicates that the PMT gain, laser power, or filter position in the current scan settings may have been changed; if the old chip is also similarly low, it may be due to scanner laser attenuation or a dirty optical path, and you need to contact the equipment engineer for calibration. At the same time, check the scanner self-test report and the last calibration date. Do not rush to redo the 8 chips; first confirm the scanning-side problem, otherwise replacing the chips may also yield the same low signal.

Don't discard or over-process the data yet. When contacting the manufacturer's technical support, have these materials ready: scanner model and PMT settings, scan date, barcodes or lot numbers of the 8 arrays, hybridization buffer and wash solution lot numbers, absolute values and background values of internal control signals, raw .tif and .xys or .CEL files. Manufacturers usually perform a technical evaluation first, and some may offer free re-runs or credits, but this depends on the contract and QC terms, so do not assume compensation is guaranteed. Before submitting a request, it is best to confirm whether the batch of arrays is within its shelf life and whether the full set of original manufacturer reagents was used. For data analysis, if there is only an overall decrease in signal while background remains stable, quantile normalization or RMA can be used to check whether between-group differences remain robust; if the detection rate drops markedly, re-running is recommended rather than forcing an interpretation.

All 8 arrays in the same batch are low, which may also be due to array batch differences or changes in the lot of hybridization buffer or wash solution. Low internal reference gene signal is usually not a sample RNA issue, because the RIN is above 7.5 and the protocol was followed. It is recommended to compare the barcode and lot number of this batch of arrays with the previous batch to see whether the lot was changed; at the same time, check whether the hybridization oven temperature is accurate, whether the hybridization time has been shortened, and whether the shaker speed is normal. Keep good logs of hybridization and washing. If the array lot is suspected, you can request the factory QC data for that lot from the manufacturer and compare the signal distribution of the internal reference probes. Generally, the manufacturer will require you to provide the original scan files first, and then determine whether the conditions for replacement or rerun are met.