Introduction

[TanyaS08]

I've pushed a first pass at the introduction. Currently the storyline is: Networks have these properties and a great mathematical framework -> We can and should decode them to make predictions -> Phylogenetic signal (TODO though) -> Networks are difficult to sample we should predict -> We need to be able to predict networks when there is no info here is our method/workflow

I don't think it's the most linear process and it might make sense to start with 'Networks are difficult to sample we should predict' and then the rest as is?