Source Data

library(dplyr)
library(qSIP2)
packageVersion("qSIP2")
#> [1] '0.22.5'

Source Material and Metadata

In a SIP experiment, the “source material” are the subjects that you are running experiments with (e.g. a culture tube or a plant root). For qSIP the source material would then be each DNA extraction that is loaded into its own column for isopycnic centrifugation. The “source data” is the highest level of metadata with a row corresponding to each original experimental or source material object. Because each source is fractionated into samples, it will have a one-to-many relationship with the “sample data” (see vignette("sample_data")).

There are a few required columns for valid source data including a unique ID (the source_mat_id), some measure of quantitative abundance for the source material (either total DNA or qPCR copies), and an isotope and substrate designation (the isotope and isotopolog, respectively). Ideally, the substrate should be a standardized compound ID (e.g. PubChem 6137 for L-methionine), but for qSIP2 is can just be descriptive text like “methionine”. In addition to the required columns, the source data can contain as many other ancillary columns as necessary to describe the experiment. These additional columns might contain important experiment-specific metadata that you will use to group and subset your source material in the qSIP workflow. But, they can also be further details that you might not need for qSIP, but it may make sense to just keep everything included if they’re already in your .txt. or excel file.

An example source dataframe is included in the qSIP2 package called example_source_df. It includes a 13C glucose addition study with two different moisture treatments (“normal” and “drought”) in quadruplicate, with one only in triplicate. Each experiment contains both unlabeled 12C and labeled 13C source material, but you may have an experiment where different 13C treatments share the same 12C source material. For example, a split experiment where you have one 12C data set that is split into many experimental conditions each with a different isotopolog.

The first few rows of example_source_df

source	total_copies_per_g	total_dna	Isotope	Moisture	isotopolog
S149	34838665	74.46539	12C	Normal	glucose
S150	53528072	109.01522	12C	Normal	glucose
S151	95774992	182.16852	12C	Normal	glucose
S152	9126192	23.68963	12C	Normal	glucose
S161	41744046	67.62552	12C	Drought	glucose
S162	49402713	94.21217	12C	Drought	glucose

qSIP2 Source Data Object

Once the dataframe is ready with at least the three required columns (source_mat_id, isotope and isotopolog), the next step is to convert it to a qsip_source_data object. This is one of the main qSIP2 objects to hold and validate the data.

source_object <- qsip_source_data(example_source_df,
  isotope = "Isotope",
  isotopolog = "isotopolog",
  source_mat_id = "source"
)

class(source_object)
#> [1] "qSIP2::qsip_source_data" "S7_object"

While these three columns are required for the EAF workflow, there are additional columns required for the growth workflow (timepoint, total_abundance and volume). These can remain empty/unassigned for this vignette, and will be detailed in a forthcoming growth workflow vignette.

Note, your column names in the dataframe don’t have to specifically be the required column names, so no need to edit your original table headers if they don’t match. For example, if you’re isotopolog column is titled “substrate”, it isn’t necessary to rename your column. If your column names are already standardized names, then there is no need to assign while creating the object. For example, the isotopolog column is already title “isotopolog”, so if it is omitted from the object creation then the column will still be identified and used.

# this will still work even though the isotopolog parameter is not assigned
qsip_source_data(example_source_df,
  isotope = "Isotope",
  source_mat_id = "source"
)

Structure of qsip_source_data

While this object is not meant to be inspected or worked with outside of qSIP2 functions, a quick glimpse() can show the structure of it.

glimpse(source_object)
#> <qSIP2::qsip_source_data>
#>  @ data           : tibble [15 × 6] (S3: tbl_df/tbl/data.frame)
#>  $ isotope           : chr [1:15] "12C" "12C" "12C" "12C" ...
#>  $ isotopolog        : chr [1:15] "glucose" "glucose" "glucose" "glucose" ...
#>  $ source_mat_id     : chr [1:15] "S149" "S150" "S151" "S152" ...
#>  $ total_copies_per_g: num [1:15] 34838665 53528072 95774992 9126192 41744046 ...
#>  $ total_dna         : num [1:15] 74.5 109 182.2 23.7 67.6 ...
#>  $ Moisture          : chr [1:15] "Normal" "Normal" "Normal" "Normal" ...
#>  @ isotope        : chr "Isotope"
#>  @ isotopolog     : chr "isotopolog"
#>  @ source_mat_id  : chr "source"
#>  @ timepoint      : chr "NULL"
#>  @ total_abundance: chr "NULL"
#>  @ volume         : chr "NULL"

The original dataframe is contained in the @data slot, however, some column names have been modified to the standard names, while keeping a record of the original names in the corresponding slots.

Column name differences

Original Names	qSIP Names	Original Name Slot
source	source_mat_id	@source_mat_id
Isotope	isotope	@isotope
substrate	isotopolog	@isotopolog

To get the dataframe back out of the qsip_source_data object you can use the get_dataframe() method with original_headers set to TRUE or FALSE, depending on your needs. But, note that the columns may be in a different order than the dataframe you started with.

get_dataframe(source_object, original_headers = T)

Isotope	isotopolog	source	total_copies_per_g	total_dna	Moisture
12C	glucose	S149	34838665	74.46539	Normal
12C	glucose	S150	53528072	109.01522	Normal
12C	glucose	S151	95774992	182.16852	Normal
12C	glucose	S152	9126192	23.68963	Normal
12C	glucose	S161	41744046	67.62552	Drought
12C	glucose	S162	49402713	94.21217	Drought
12C	glucose	S163	47777726	87.82524	Drought
12C	glucose	S164	48734282	75.97274	Drought
13C	glucose	S178	62964478	73.89526	Normal
13C	glucose	S179	49475460	68.65182	Normal
13C	glucose	S180	51720787	81.36874	Normal
13C	glucose	S200	59426155	71.19377	Drought
13C	glucose	S201	56379702	73.78225	Drought
13C	glucose	S202	42562198	108.11436	Drought
13C	glucose	S203	49914369	80.48608	Drought

Validation of qsip_source_data

While constructing a qsip_source_data object there are a few validation checks that are performed. For now, the only checks are that the source_mat_id is unique for each row, and that the isotope field is an appropriate value. This doesn’t just mean it is a value that makes sense, but also that it is one of the isotopes that qSIP2 knows how to calculate atom fraction values from. This is currently limited to 12C/13C, 14N/15N and 16O/18O. There are some “non-isotopic” names allowed as well for source material that might be unfractionated. These additional options are “bulk”, “unfractionated”, “T0”, “time0”, “Time0”, and are added as exceptions in the validate_isotopes() helper function.

# artificially doubling the rows will give an error from duplicate source_mat_ids
example_source_df |>
  rbind(example_source_df) |>
  qsip_source_data(
    isotope = "Isotope",
    isotopolog = "isotopolog",
    source_mat_id = "source"
  )
#> Error:
#> ! some source_mat_ids are duplicated

One benefit of the validation steps being embedded in the object itself is that these validations are automatically run when the object is modified. This makes it impossible to modify the data later to an invalid object, e.g. changing an isotope to an invalid choice.

source_object@data$isotope <- "13G"
#> invalid isotope found: 13G
#> Error:
#> ! Please fix the isotope names and try again

MISIP

While qSIP standards are part of the MISIP¹ standards, the qSIP2 package is a little less stringent. This means your valid qSIP2 object might not be valid for a MISIP submission. At the source data level this is primarily through the difference between how the isotope data is coded, plus the addition of another isotopolog_label column.

qSIP2 has functions to convert between these two types. add_isotoplog_label() makes a MISIP version of the source data, and remove_isotopolog_label() converts it back to a qSIP2 compatible version. Two things are changed when running add_isotoplog_label() - 1) the isotopolog_label column is added and is populated with either “isotopically labeled” or “natural abundance” for heavy and light isotopes, respectively, and 2) the isotope column gets modified to be only the heavy isotope (e.g. all “12C” entries become “13C”).

Note, these functions are run on the source dataframe rather than on the qsip_source_data object.

Adding the isotopolog_label column

example_source_df |>
  add_isotopolog_label(isotope = "Isotope")

A ’MISIPified version of example_source_df

source	total_copies_per_g	total_dna	isotope	isotopolog_label	Moisture	isotopolog
S151	95774992	182.16852	13C	natural abundance	Normal	glucose
S178	62964478	73.89526	13C	isotopically labeled	Normal	glucose
S200	59426155	71.19377	13C	isotopically labeled	Drought	glucose
S201	56379702	73.78225	13C	isotopically labeled	Drought	glucose
S150	53528072	109.01522	13C	natural abundance	Normal	glucose
S180	51720787	81.36874	13C	isotopically labeled	Normal	glucose

Now, the Isotope column has been renamed to isotope to satisfy MISIP standards, and all values have been replaced with the heavy isotope.

Count of isotope types in example_source_df_MISIP

isotope	n
13C	15

And the designation for whether the source material was the “light” or “heavy” version of the isotope has now been transferred to the isotopolog_label column.

Count of isotopolog_label types in example_source_df_MISIP

isotopolog_label	n
isotopically labeled	7
natural abundance	8

Removing the isotopolog_label column

This change can be reverted with the remove_isotopolog_label() function.

example_source_df |>
  add_isotopolog_label(isotope = "Isotope") |>
  remove_isotopolog_label()

example_source_df_MISIP converted back

source	total_copies_per_g	total_dna	isotope	Moisture	isotopolog
S149	34838665	74.46539	12C	Normal	glucose
S150	53528072	109.01522	12C	Normal	glucose
S151	95774992	182.16852	12C	Normal	glucose
S152	9126192	23.68963	12C	Normal	glucose
S161	41744046	67.62552	12C	Drought	glucose
S162	49402713	94.21217	12C	Drought	glucose
S163	47777726	87.82524	12C	Drought	glucose
S164	48734282	75.97274	12C	Drought	glucose
S178	62964478	73.89526	13C	Normal	glucose
S179	49475460	68.65182	13C	Normal	glucose
S180	51720787	81.36874	13C	Normal	glucose
S200	59426155	71.19377	13C	Drought	glucose
S201	56379702	73.78225	13C	Drought	glucose
S202	42562198	108.11436	13C	Drought	glucose
S203	49914369	80.48608	13C	Drought	glucose

Note, the original is not exactly preserved as the original Isotope column has the MISIP standard isotope name retained.