download abstracts

I am using the package 'RISmed'.


Key steps:

Get the Pubmed IDs. 

This requires the function EUtilsSummary().
Here is the code to find Pubmed IDs for publications by C. Pepper over the last five years:

res <- EUtilsSummary("Pepper_C[au]", type='esearch', db='pubmed', mindate=2010, maxdate=2015)

This creates the object res which is a list of Pubmed IDs.
It called a "Formal class EUtilsSummary".
To find out a bit more about the object the following commands are useful:

> str(res)
Formal class 'EUtilsSummary' [package "RISmed"] with 6 slots
  ..@ db              : chr "pubmed"
  ..@ count           : num 77
  ..@ retmax          : num 77
  ..@ retstart        : num 0
  ..@ PMID            : chr [1:77] "25921075" "25897547" "25806611" "25752197" ...
  ..@ querytranslation: chr "Pepper_C[au] AND 2010[EDAT] : 2015[EDAT]"
> mode(res)
[1] "S4"
> summary(res)
Query:
Pepper_C[au] AND 2010[EDAT] : 2015[EDAT] 

Result count:  77
> 

So usefully this tells us about the search we have done and the number of results. The results are in the @PMID part of the object. The object is of class S4. 

You can extract the individual PMIDs like this:
> res@PMID[1]
[1] "25921075"
> res@PMID[2]
[1] "25897547"
> res@PMID[3]
[1] "25806611"

Useful!!!

Downloading the abstracts

This res object can be used to download the abstracts using the EUtilsGet() function.

fetch <- EUtilsGet(res)

This returns an object called fetch. 
This is called "Large Medline (618.6 Kb)"
It contains lots of data including the abstracts. 

Extracting the abstracts

The abstracts can be extracted with the function AbstractText(). 
abs <- AbstractText(fetch)
Now we have 77 abstracts in a vector (I think).

> str(abs)
 chr [1:77] "OBJECTIVES: Ganglioneuroblastomas represent a histological subgroup of the rare neuroblastic tumours with intermediate malignan"| __truncated__ ...

> mode(abs)
[1] "character"

> abs[1]
[1] "OBJECTIVES: Ganglioneuroblastomas represent a histological subgroup of the rare neuroblastic tumours with intermediate malignant potential arising from neural crest progenitor cells of sympathetic nerves. Diagnosis can often be difficult based on imaging alone. We describe 4 cases of children presenting with a solitary neck mass with histology ultimately revealing ganglioneuroblastoma.METHODS: A retrospective case note review was carried out of all patients with cervical ganglioneuroblastoma seen at Great Ormond Street Hospital, UK.RESULTS: Mean age at presentation was 5 years. Based on imaging, the initial diagnoses for three of the cases were: lymphatic malformation, carotid body tumour, paraganglioma, respectively, whilst the remaining case had an immediate incisional biopsy revealing the correct diagnosis. All cases were managed by surgical excision with no evidence of recurrence after a median follow up of 6 years.CONCLUSION: Otolaryngologists should be aware of ganglioneuroblastoma when establishing the differential diagnosis of a child presenting with a neck mass. Biopsy is recommended as the gold standard investigation to avoid an incorrect diagnosis."
> 

Great. Good so far....