The enrichment of signal values near the TSS on the sense and antisense strands #1403

My RNA data was generated using a dUTP-based library preparation method, aligned using STAR, and then processed with bamCoverage using the --filterRNAstrand parameter to generate forward and reverse strand bigWig files. The --scaleFactor > 0. And input these bigWig files into computeMatrix reference-point with --referencePoint TSS to compute the signal. The resulting profile plot shows signal accumulation to the right of the TSS for both strands. However, in other published plots, the signal for the antisense strand accumulates to the left of the TSS. Am I missing some parameter or some data processing? I also noticed a similar question raised in issue #649 , but no clear answer was provided.