The data frames that are returned by the dbparser package (using the drug_parse_all() function) do not all have the same class/type. Observe the following examples:

> get_xml_db_rows('full database.xml')

> myall <- parse_drug_all()

> class(myall$drug_targets)
[1] "tbl_df"     "tbl"        "data.frame"

> class(myall$drug_targets_actions)
[1] "data.frame"

Please ensure that all the returned data frames consistently have the same class/type:
"tbl_df" "tbl" "data.frame" (i.e. dplyr's version of data frame, tibble)

Is your feature request related to a problem? Please describe.
Instead of downloading the file and give it to the package, the package can download it directly given Drugbank user name and password

The vignette's figures require the full DrugBank database to be displayed. Since the full database would be be too big to include in the package, this prevents the vignette from being self-contained (which is a bad thing).

This issue is about coming up with another couple of EDA examples to add in the vignette that does not require the full database. The purpose of this is to avoid relying on the full DrugBank database so that the vignette may be self-contained.

When using the dbparser::parse_drug() function, the returned data frame has many variables ending with the _count suffix. In RStudio, when one is looking at the different elements of the data frame (by adding a $ sign and pressing TAB), it can be a bit annoying to find these count variables cluttering the elements list.

Proposed enhancement:
Instead of a ..._count suffix, let these variables have a count_... prefix so that they are bundled together; i.e. make use of the fact that the elements are ordered alphabetically in the list returned when pressing TAB.

Bonus:
In the list returned by the tidyverse's glimpse() function, it would be great the count elements are moved to the end of that list.

resolve any issue from package checks before submitting to CRAN

Gather all the data frames that are returned by dbparser's parse_drug_all() function, and submit to Kaggle as a dataset. This dataset is to be updated quarterly with each new release of the DrugBank dataset.

Prerequisites:

• Ask permission from DrugBank and make sure it is okay to do such a thing.
• Homework: Can we contribute the dataset as a team (Dainanahan)?
• Challenge: The DrugBank data is >1GB in size. How to upload such a dataset?
  • Suggested Solution: Rent an online VM for a day. From that VM, download and parse DrugBank XML, then upload to Kaggle.
• The dbparser website has to be finalized.
  • Meaning of the word, 'finalized', to be discussed and specified later, but I am mainly referring to tutorials, usage examples, blogs and inspiration from our previously submitted DataCamp proposal.
• Prior to contributing the dataset, there needs to be enough tutorial/blog material for Kagglers to know what to do with the data (what analyses to perform? what interesting questions could be answered with the dataset? perhaps contribute a Kaggle kernel of our own immediately after the dataset is posted online? etc.)
• Homework: We have to make sure that the names of the variables (columns) of the data frames (returned by dbparser) are acceptable and in the way that we want them (i.e. self-explanatory without much needed elaboration from our side).
  • This will require a good look at the data frames' column names.
  • The parent_key variable in particular will be revisited (e.g. wherever parent_key refers to a drug, change column name to drugbank_id?).

http://download.bio2rdf.org/release/3/drugbank/drugbank.html

GO_Classifiers_Polypeptide_Target_Drug is returned twice and drug_carriers_polypeptides_go_classifiers is not following name standard

dbparser's parse_drug_all() function returns all the parsed data in the form of a list of 72 data frames. All the data frames have the prefix drug_ except two of them:

• experimental_properties: please change to drug_experimental_properties
• classificationsparse_drug_classifications: please change to drug_classification

Additionally, please make sure that the names of the data frames returned by the individual parsing functions, parse_drug_experimental_properties() and parse_drug_classifications(), are also correct.

Finally, I believe that parse_drug_classifications() should be parse_drug_classification() (i.e. remove the 's') to correctly correspond with the <classification> tag in the XML (in the same way that parse_drug() corresponds to the <drug> tag).

Using dbparser's parse_drug_all() function, the user is able to obtain the entire data residing within the XML file containing the DrugBank database. The parsed data is in the form of a list of data frames. This is fine in the case where the user is interested in doing an analysis that involves the entire data.

This issue is about the case when the user is only interested in a specific drug (or group of drugs). Let's refer to the list of data frames (returned by parse_drug_all()) as dblist.

A function called filter_drugs() is to be added that would be used as follows:

• First, the user would programmatically filter just the drugs data frame from dblist using whatever criteria they wish (e.g. via the dplyr::filter() function).
• After the user is satisfied with the set of drugs in dblist$drugs, dblist is then passed to filter_drugs().
• As a result, all the other data frames in dblist would be filtered by the set of drugs in dblist$drugs.
• dblist is then returned to the user after having been filtered.

Bonus: The function is preferred to have error-handling and/or warning messages for when unexpected things happen (e.g. a data frame from dblist is missing).

update tutorial on how to reproduce it

[1] "Parsed Synonyms_Polypeptide_Carrier_Drug, 30/74"
Error in xmlChildren(p[["pfams"]])[[1]] : subscript out of bounds

Example of branding on RStudio blog
https://blog.rstudio.com/2018/01/29/sparklyr-0-7/

The data frame returned by parse_drug() has, as the main drug ID, the column named 'primary_key'. However, when the drug IDs appear in data frames returned by other parse*() functions, it is typically under a column named 'parent_key'.

Here, I am suggesting to modify the 'primary_key' column name in the parse_drug() data frame to 'parent_key' so that it would match the other data frames. This would mainly be convenient in joining operations between data frames that will be done quiet frequently in the future.

The following warning is displayed upon using any parse*() function:
"In bind_rows_(x, .id) : binding character and factor vector, coercing into character vector"

It is my impression that CRAN wants there to be no such warning messages for typical/standard/default usage of any of dbparser's functions. The above message shows up when running the function on the full DrugBank database XML.

When attempting to read a non-existing XML file using the get_xml_db_rows('<FILENAME.xml>') function of the dbparser package, an error is given (which makes sense). However, the error produced in this scenario is:

Error: XML content does not seem to be XML: '<FILENAME.xml>'

The issue here is that the error message is incorrect since the actual problem is that the file does not exist in the first place (and not that the file exists and its contents are non-XML). The error message is currently misleading and is supposed to be something like the one below instead:

Could not find the file: "<FILENAME.xml>". Please ensure that the file name is entered correctly and that it exists at the specified location.

Naturally, code that checks for the file's existence will need to be added, and if indeed the file doesn't exist, produce the error message above.

Is your feature request related to a problem? Please describe.
Extracting SMILES and similar attributes from Small Molecule drugs

Describe the solution you'd like
Under Drugs element:
<property> <kind>SMILES</kind> <value>CC[C@H](C)[C@H](NC(=O)[C@H](CCC(O)=O)NC(=O)[C@H](CCC(O)=O)NC(=O)[C@H](CC1=CC=CC=C1)NC(=O)[C@H](CC(O)=O)NC(=O)CNC(=O)[C@H](CC(N)=O)NC(=O)CNC(=O)CNC(=O)CNC(=O)CNC(=O)[C@@H]1CCCN1C(=O)[C@H](CCCNC(N)=N)NC(=O)[C@@H]1CCCN1C(=O)[C@H](N)CC1=CC=CC=C1)C(=O)N1CCC[C@H]1C(=O)N[C@@H](CCC(O)=O)C(=O)N[C@@H](CCC(O)=O)C(=O)N[C@@H](CC1=CC=C(O)C=C1)C(=O)N[C@@H](CC(C)C)C(O)=O</value> <source>ChemAxon</source> </property>

Observation
There are many duplicate records in the data frames returned by parse_drug_all() that start with the following prefixes:

• drug_targets_
• drug_enzymes_
• drug_transporters_
• drug_carriers_

How to Reproduce Observation
Observe the following examples:

> library(dbparser)
> get_xml_db_rows('full database.xml')
> my_all <- parse_drug_all()
>
>
>
> my_all$drug_targets_polypeptide %>% nrow
[1] 18639
> my_all$drug_targets_polypeptide %>% unique %>% nrow
[1] 4763
>
> my_all$drug_targets_polypeptides_external_identifiers %>% nrow
[1] 107715
> my_all$drug_targets_polypeptides_external_identifiers %>% unique %>% nrow
[1] 21517
>
> my_all$drug_targets_actions %>% nrow
[1] 7671
> my_all$drug_targets_actions %>% unique %>% nrow
[1] 2953
>
>
>
> my_all$drug_enzymes_actions %>% nrow
[1] 5286
> my_all$drug_enzymes_actions %>% unique %>% nrow
[1] 517
>
>
>
> my_all$drug_carriers_actions %>% nrow
[1] 251
> my_all$drug_carriers_actions %>% unique %>% nrow
[1] 85

What to Do?
For any of the data frames with the above-mentioned prefixes, please apply the unique() function to them in order to remove any duplicate rows before returning them to the user.

Suggestion
It may even be a good idea to apply the unique() function to ALL data frames produced by dbparser before returning them to the user. However, I will let you be the judge of that. :)

Add a function, e.g. parse_everything(), that parses all the data and returns the parsed data in the form of a list of dataframes.

Note that if save_table=TRUE, all the parsed data should be successfully get saved in the DB.

It would be nice if the XML file containing the DrugBank database could be parsed while still in its compressed form. Specifically, if the file ends in .zip, the file is to be uncompressed automatically before parsing.

Inspiration for this issue came from the file argument of the read_csv() function of the dplyr package.

Bonus: If dbparser is able to deal with other compressed formats (.gz, .bz2, .xz), that would be great!

Bonus: If the filename starts with http://, https://, ftp://, or ftps://, then additionally download the file prior to uncompressing and parsing.

https://www.rplumber.io/

• Need to update parse all as well, to not parse an element with an existing csv.
• Support overriding the written csv if user wants to

There are some attributes that still need to be parsed such as:

• Calculated Properties.
• average-mass
• etc...

Happy to send a PR with badges if interested!

Review should be for semantic and syntax errors

The latest commit by you seems to cause a build error, please check
https://travis-ci.org/Dainanahan/dbparser/builds/474200685

it has usage no more

The figures in the vignette look unsatisfactory. Specifically, the figures are left-centered and tiny. A quick comparison between the figures in dbparser's vignette and another example vignette from tidytext clearly shows that work is needed in dbparser's figures.

• please write package names, software names and API names in
  single quotes (e.g. 'DrugBank') in Title and Description.
• Please add an URL for 'DrugBank' in your Description text in the form
  http:... or https:...
  with angle brackets for auto-linking and no space after 'http:' and
  'https:'.
• Please replace \dontrun{} by \donttest{} or unwap the examples if they
  can be executed in less than 5 sec per Rd-file.

It would be convenient for Python users who are interested in the dbparser package if there exists a Python wrapper for it. Such a wrapper may, for example, be uploaded to Anaconda Cloud or to something that is like BioConductor but within the Python universe (e.g. BioPython).

Is your feature request related to a problem? Please describe.
Putting name standard to the retrieved datasets will make dealing with it much more easier in dependent packages and apps, please refer to DrugBank Browser

Describe the solution you'd like

• Name should be DatasetName_Parent(s)Name(s), i.e Drugs.csv, Names_Drugs.csv, etc
• Ignore references and interactions in names

Three misspelled file names were found in the dbparser package:

• dbparser/tests/testthat/
  • test_durg_main_node_parser.R
  • test_durg_all_nodes_parser.R
• dbparser/R/
  • durg_main_node_parser.R

Please, fix the file names and make sure there are no similar misspellings inside the above files.

More detailed documentation is needed

Current only option is saving the dataset into database, we could save it in user hard disk as well

Drugs have sometimes more than 3 ids

<drugbank-id primary="true">DB00006</drugbank-id>
  <drugbank-id>BTD00076</drugbank-id>
  <drugbank-id>EXPT03302</drugbank-id>
  <drugbank-id>BIOD00076</drugbank-id>
  <drugbank-id>DB02351</drugbank-id>

Need to handle any number of drugbank ids

When using the dbparser::parse_drug_targets_polypeptides() function, the returned data frame has a misspelled variable, amindo_acid_format.

Proposed Solution:
amindo_acid_format --> amino_acid_format

Bonus:
See if you can find similar typos, and fix them too!

Is your feature request related to a problem? Please describe.
When I used the generated dataframes from dbparser I had to rename its features to be displayed properly

Describe the solution you'd like
Rename features like text with a proper name, features with small caps to be Camel case

Currently, save_table is TRUE by default for all parse*() functions. Let it be FALSE by default. The change is to be reflected in all functions' documentations (i.e. the kind of help page that displays when adding a '?' before the function name in the console and pressing Enter).

The documentation page for dbparser shows up when you type the following and press Enter:
help(package = dbparser)

In the page that is displayed, the title reads:
Drug bank xml db parser and database saver

I prefer the title to be:
DrugBank Database XML Parser

An overall check of the entire documentation (including all the parse*() functions) would be welcome (for typos and clarity), but the title is enough for now.

Is your feature request related to a problem? Please describe.
These features has no use currently as they will be replaced by drugQuery package

Describe the solution you'd like
New package called drugQuery will replace these features and provide more complex and advanced queries on drug bank

Describe the bug
There are two datasets are missing in the parser salts and international-brands