offst(newSeqExpressionSet) contains NaNs and Inf values

[albptt]

Good morning,
I am trying to perform GC content normalization using the following workflow:

expr_set <- newSeqExpressionSet(as.matrix(raw_counts), featureData = as.data.frame(geneInfoHT))

expr_set <– withinLaneNormalization(expr_set, "gcContent", which = "upper", offset = TRUE)

expr_set <- betweenLaneNormalization(expr_set, which = "upper", offset = TRUE)

normalized_counts <-  log(raw_counts[rownames(expr_set),] + .1) + offst(expr_set)

However, I notice that even if the raw_counts do not contain neither NaNs nor Inf values, the offset returned by the withinLaneNormalization function contains 4290 NaNs and 1347 Inf values.

Might these values be expected?

Thank you for the help!