/
pfocr_fetch.R
210 lines (181 loc) · 7.61 KB
/
pfocr_fetch.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
## fetch pathway figure PMCIDs from NCBI
## NOTE: query qualifier for figure captions [CAPT] is clearly broken and only hits on a fraction of caption titles.
## the "imagesdocsum" report type does a better job of actually searching captions, e.g.:
# https://www.ncbi.nlm.nih.gov/pmc/?term=(signaling+pathway)+AND+(2019+[pdat])&report=imagesdocsum&dispmax=100
## (11349 hits with "signaling pathway" in every caption title or caption body)
# https://www.ncbi.nlm.nih.gov/pmc/?term=(signaling+pathway[CAPT])+AND+(2019+[pdat])&report=imagesdocsum&dispmax=100
## (244 hits with "signaling pathway" ONLY in caption titles)
# https://www.ncbi.nlm.nih.gov/pmc/?term=(signaling+pathway[CAPT])+AND+(2019+[pdat])
## (2775 hits when "report=imagesdocsum" is excluded)
## NOTE: the imagesdocsum" report is not supported by NCBI's eutils, so we'll have to go with HTML scraping.
## The pagination of pmc output is not apparent, however...
## Example queries for what is possible
# https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=pubmed&term=asthma[mesh]+AND+leukotrienes[mesh]+AND+2009[pdat]&usehistory=y&retmax=500&retStart=0
# https://www.ncbi.nlm.nih.gov/pmc/?term=signaling+pathway+AND+2018+[pdat]&report=imagesdocsum&dispmax=100
# https://www.ncbi.nlm.nih.gov/pmc/?term=((((((((((signaling+pathway)+OR+regulatory+pathway)+OR+disease+pathway)+OR+drug+pathway)+OR+metabolic+pathway)+OR+biosynthetic+pathway)+OR+synthesis+pathway)+OR+cancer+pathway)+OR+response+pathway)+OR+cycle+pathway)+AND+(\%222019/01/01\%22[PUBDATE]+%3A+\%223000\%22[PUBDATE])&report=imagesdocsum&dispmax=100#
## Network query:
# https://www.ncbi.nlm.nih.gov/pmc/?term=((network)+OR+PPI)+AND+(%222019/01/01%22[PUBDATE]+%3A+%223000%22[PUBDATE])&report=imagesdocsum&dispmax=100
##################
## QUERY BUILDER
##################
## Pathway types:
query.terms <- c("signaling+pathway",
"signalling+pathway",
"regulatory+pathway",
"disease+pathway",
"drug+pathway",
"metabolic+pathway",
"biosynthetic+pathway",
"synthesis+pathway",
"cancer+pathway",
"response+pathway",
"cycle+pathway"
)
query.date.from <- "1995/01/01"
query.date.to <- "3000/01/01"
term <- paste0("term=",paste(rep("(",length(query.terms)),collapse = ""),paste(lapply(query.terms, function(x){
paste0(x,")")
}), collapse = "+OR+"),
'+AND+("',query.date.from,'"[PUBDATE]+%3A+"',query.date.to,'"[PUBDATE])')
query.url <- paste0("https://www.ncbi.nlm.nih.gov/pmc/?",
term,
"&report=imagesdocsum",
"&dispmax=100")
################
## PMC SCRAPER
################
library(conflicted)
library(RSelenium)
library(rvest)
library(xml2)
library(tidyverse)
conflict_prefer("filter", "dplyr")
conflict_prefer("select", "dplyr")
conflict_prefer("mutate", "dplyr")
# set dir for saving results as tsv
setwd("/git/wikipathways/pathway-figure-ocr/20200214")
cat(query.url, file="query.txt")
write.table(data.frame("figureid","pmcid", "filename", "fignumber", "figtitle", "papertitle", "figcaption", "figlink", "reftext"),
file = "pmc.df.all.tsv",
append = FALSE,
sep = '\t',
quote = FALSE,
col.names = FALSE,
row.names = FALSE)
# launch Docker.app
# run this line in Terminal: docker run -d -p 4445:4444 selenium/standalone-firefox:2.53.0
remDr <- remoteDriver(
remoteServerAddr = "localhost",
port = 4445L
)
remDr$open()
## go to query result
remDr$navigate(query.url)
# confirm you got there
remDr$screenshot(display = TRUE)
## Collect all pages!
df.all <- data.frame(figid=character(),pmcid=character(), filename=character(),
number=character(), figtitle=character(), papertitle=character(),
caption=character(), figlink=character(), reftext=character())
page.count <- xml2::read_html(remDr$getPageSource()[[1]]) %>%
rvest::html_nodes(".title_and_pager") %>%
rvest::html_node(".pagination") %>%
rvest::html_nodes("a") %>%
rvest::html_attr("page")
page.count <- as.integer(page.count[4])
for (i in 1:page.count){
## Parse page
page.source <- xml2::read_html(remDr$getPageSource()[[1]])
filename <- page.source %>%
rvest::html_nodes(".rprt_img") %>%
rvest::html_node("img") %>%
rvest::html_attr("src-large") %>%
str_match("bin/(.*\\.jpg)") %>%
as.data.frame() %>%
select(2) %>%
as.matrix() %>%
as.character()
number <- page.source %>%
rvest::html_nodes(".rprt_img") %>%
rvest::html_node("img") %>%
rvest::html_attr("alt")
titles <- page.source %>%
rvest::html_nodes(".rprt_img") %>%
rvest::html_node(xpath='..') %>%
rvest::html_node(".rprt_cont") %>%
rvest::html_node(".title") %>%
rvest::html_text() %>%
str_split("\\s+From: ", simplify = TRUE)
papertitle <- titles[,2] %>%
str_trim()
caption <- page.source %>%
rvest::html_nodes(".rprt_img") %>%
rvest::html_node(xpath='..') %>%
rvest::html_node(".rprt_cont") %>%
rvest::html_node(".supp") %>%
rvest::html_text()
figlink <- page.source %>%
rvest::html_nodes(".rprt_img") %>%
rvest::html_attr("image-link")
reftext <- page.source %>%
rvest::html_nodes(".rprt_img") %>%
rvest::html_node(xpath='..') %>%
rvest::html_node(".rprt_cont") %>%
rvest::html_node(".aux") %>%
rvest::html_text() %>%
str_remove(fixed("CitationFull text"))
pmcid <- page.source %>%
rvest::html_nodes(".rprt_img") %>%
rvest::html_node(xpath='..') %>%
rvest::html_node(".rprt_cont") %>%
rvest::html_node(".title") %>%
rvest::html_node("a") %>%
rvest::html_attr("href") %>%
str_match("PMC\\d+") %>%
as.character()
## Extract best figure title from analysis of provided, number, title and caption
temp.df <- data.frame(n=number, t=titles[,1], c=caption, stringsAsFactors = FALSE) %>%
mutate(t=str_trim(str_remove(t,fixed(as.character(number)))))
temp.df <- temp.df %>%
mutate(c=if_else(is.na(c),t,c)) %>%
mutate(t=str_trim(str_remove(t,"\\.$"))) %>%
mutate(t=if_else(t=="",c,t)) %>%
mutate(t=if_else(!is.na(str_match(t,"\\. .*")),str_remove(t,"\\. .*"),t)) %>%
mutate(t=str_trim(str_remove(t,"\\.+$"))) %>%
mutate(t=str_trim(str_remove(t,"^\\."))) %>%
mutate(c=str_trim(str_replace(c,"\\.\\.","\\."))) %>%
mutate(n=str_trim(str_replace(n,"\\.$","")))
number <- as.character(temp.df[,1])
figtitle <- as.character(temp.df[,2])
caption <- as.character(temp.df[,3])
## Prepare df and write to R.object and tsv
df <- data.frame(pmcid, filename, number, figtitle, papertitle, caption, figlink, reftext) %>%
mutate(figid=paste(pmcid,filename, sep = "__")) %>%
select(figid,pmcid, filename, number, figtitle, papertitle, caption, figlink, reftext)
df.all <- rbind(df.all, df)
write.table(df,
file = "pmc.df.all.tsv",
append = TRUE,
sep = '\t',
quote = FALSE,
col.names = FALSE,
row.names = FALSE,
fileEncoding = "UTF-8")
if (i < page.count-1){
next.page.button <- remDr$findElement(using = "xpath", "//*[@class='active page_link next']")
next.page.button$clickElement()
#remDr$screenshot(display = TRUE)
}
print(paste(i,"of",page.count))
}
#If error, go back a page (next line) and enter last page number as new start in for loop
#remDr$goBack()
## At the end of the day...
df.all<-unique(df.all)
saveRDS(df.all, file = "pmc.df.all.rds")
#df.all <- readRDS("pmc.df.all.rds")
## Close up shop
remDr$closeall()
# In Terminal:
# docker ps ## note CONTAINER ID
# docker stop <CONTAINER ID>