How to perform replacements containing square brackets with string_replace()?

I have a vector with two elements:

xx <- c("K[Phenylisocyanate (N-term)]SAPATGGVKKPHR", "A[Phenylisocyanate (N-term)]SAPATGGVKKPHR")

I would like to replace "[Phenylisocyanate (N-term)]" with "[Phenylisocyanate_Propionyl (Nterm)]", but only if the part to be replaced ("[Phenylisocyanate (N-term)]") is preceded by "K", i.e. in the first element only

So I would like to get c("K[Phenylisocyanate_Propionyl (Nterm)]SAPATGGVKKPHR", "A[Phenylisocyanate (N-term)]SAPATGGVKKPHR").

Performing this replacement with

str_replace (xx,  "K[Phenylisocyanate (N-term)]", "K[Phenylisocyanate_Propionyl (Nterm)]")

results in very unexpected behaviour:

[1] "K[Phenylisocyanate_Propionyl (Nterm)]Phenylisocyanate (N-term)]SAPATGGVKKPHR"

[2] "A[Phenylisocyanate (N-term)]SAPATGGVKK[Phenylisocyanate_Propionyl (Nterm)]HR"

could you perhaps explain how I should be writing this replacement and why I get this bizzare result?

>Solution :

The reason for the result above is that [ and ] are metacharacters in regular expressions, they mean ‘match any character that appears inside the square brackets’.

To get around this, you can use the fixed() function to tell R that the pattern is not a regular expression.

str_replace(xx,  fixed("K[Phenylisocyanate (N-term)]"), "K[Phenylisocyanate_Propionyl (Nterm)]")