sequence bias modeling understanding

[netpumber]

Hi.

I'am new at NGS , RNA-seq and now i'm reading an article about the expression units in an RNA-seq experiment. The first method that it talks about is "Counts".

So (as i understood) the number of reads that will be aligned in the reference gene is depended on 1) the total number of DNA fragments and 2) on the effective length of the feature (e.g reference gene). Everything is ok until here.

Then author of the article writes:

"If the abundance estimation method you’re using incorporates sequence bias modeling (such as eXpress or Cufflinks), the bias is often incorporated into the effective length by making the feature shorter or longer depending on the effect of the bias."

As i told before i'm new and could not understand the meaning of the "sequence bias modeling". Does this mean that user defines a specific length (something like threshold) for the reference gene ?

Any hint or example for understanding that method and why is not so good to use, is welcome.

Thank you very much.