Input File Format

Our algorithm is really powerful. However, it still has not possessed human intelligence yet. This unfortunately means that the input files need to conform with some certain format.

In this section, we will take a deep dive into the read_file function provided in the data_curation module of pMTnet_Omni_Document package.

File Format

Although the default format that read_file expects is .csv, since under the hood, we use the read_csv function from the pandas package, read_file accepts inputs of major file formats, as long as the corresponding sep argument is supplied.

Acceptable File Format

File Format

sep

.csv

,

.txt

User defined. Usually , or \t

.tsv

\t

Column Names

As methods of pMTnet Omni manipulates the input dataframe based on its column names, harmonizing column names is necessary. Therefore, when reading the user input, read_file will first attempt to find the following column names.

Note

Details on the data format will be explained in Data Format section.

Column Name

Name

Meaning

Mandatory

Note

va

The name of the Alpha chain for the V segment

No

At least one of va and vaseq needs to be supplied

vaseq

The actual sequence of amino acids of va

No

At least one of va and vaseq needs to be supplied

vb

The name of the Beta chain for the V segment

No

At least one of vb and vbseq needs to be supplied

vbseq

The actual sequence of amino acids of vb

No

At least one of vb and vbseq needs to be supplied

cdr3a

The sequence of amino acids for the CDR3 region on the Alpha chain

Yes

cdr3b

The sequence of amino acids for the CDR3 region on the Beta chain

Yes

peptide

The sequence of amino acids presented by the MHC

Yes

mhc

The name(s) of the MHC

No

At least one of mhc and mhcseq needs to be supplied

mhcseq

The sequence(s) of amino acids of the corresponding mhc

No

At least one of mhc and mhcseq needs to be supplied

tcr_species

Species of the TCR

Yes

human or mouse

pmhc_species

Species for the peptide-MHC

Yes

human or mouse

If read_file can not match these names exactly, it will try modifying the found column names in the user input to match the names. Specifically, it will change the names to lower cases, strip all white spaces, and remove special characters like *, _, +, -. See the following for some acceptable input examples as well as some unacceptable input examples.

Example Input

Input

Acceptable

V_A seq

Yes

pMHC SPECIES

Yes

V Alpha

No

Epitope

No

Data Format

We understand that the nomenclatures are not always consistent in the field of biology. Albeit being conceptually viable, it would be too overwhelming for us to take all commonly-used naming conventions into consideration.

In this section, the required format for all fields in the user input will be elaborated upon.

Amino Acids Sequences

VA, VB

MHC