Add info about fasterq-dump (#36)

docs/main_text.md

@@ -124,7 +124,9 @@ Some analyses around this platform detection procedure can be found in [this rep
 ![rna-seq-pipeline](https://user-images.githubusercontent.com/15315514/44549339-c86fd680-a6ee-11e8-8d62-419ae7f10a94.png)
 
 We use [Salmon](https://combine-lab.github.io/salmon/) and [tximport](https://bioconductor.org/packages/release/bioc/html/tximport.html) to process all RNA-seq data in refine.bio.
-We obtain fastq files run on our [supported short-read platforms](https://github.com/AlexsLemonade/refinebio/blob/dev/config/supported_rnaseq_platforms.txt) from Sequence Read Archive. 
+We obtain sra files run on our [supported short-read platforms](https://github.com/AlexsLemonade/refinebio/blob/dev/config/supported_rnaseq_platforms.txt) from NCBI Sequence Read Archive and use [`fasterq-dump`](https://github.com/ncbi/sra-tools/wiki/HowTo:-fasterq-dump) (with default behavior) to obtain fastq files for use with Salmon.
+Note that any unmated reads from paired experiments are discarded.
+
 We use the library strategy and library source metadata fields to identify RNA-seq experiments.
 It's possible that experiments that are inappropriate for use with Salmon will still appear in refine.bio (e.g., long-read platforms that are labeled incorrectly in the source repository).
 If you find an experiment that you believe is inappropriate for use with Salmon, please [file an issue on GitHub](https://github.com/AlexsLemonade/refinebio/issues) so that we can resolve it.