Translate input cDNA to Amino Acid Sequence (Front-end Service?) - 3 Input Boxes

[David-Bianchi]

As a user, I would like a cDNA input to the tool to be able to be translated directly to the corresponding amino acid sequence (by default using the standard genetic code translation table, ID=1), or by another translation table if it is specified.

This could allow for a type of input such as what is shown below:

---

Sharepoint Description from Nilmani of Input File Definitions: https://uillinoisedu-my.sharepoint.com/:p:/r/personal/nsingh16_illinois_edu/_layouts/15/Doc.aspx?sourcedoc=%7B69146A71-68B5-47AB-8190-249901664535%7D&file=Site_Directed_Mutagenesis_NCSA.pptx&action=edit&mobileredirect=true

Design Depiction:

List of ACs:

AC1. Translates ORF to amino acid sequence
AC2. uses the standard translation table (ID=1) by default
AC3. Allows for other translation tables to be used.
We do not need to maintain AC4
AC4. The cDNA must start with ATG (or a relevant start codon for the Translation Table in Use)
AC5. No stop codon in reading frame
AC6. The cDNA contains the complete ORF, with also entry 50 bp upstream and 50 bp downstream minimum (See Image Example Below for 3-textbox input method)
AC7. Refer to cDNA as ORF throughout the back-end and front-end interface.
AC8. All bases ACTG "Non-canonical bases entered, please enter sequence containing only ACTG"

Future ACs:
F-AC1. Support for FASTA/SnapGene Input
F-AC2. The end-user will have to label cDNA/ORF in these files.

Other resources:

1. We'll need to validate that the input is cDNA on the front-end (we may have existing code to do this from other projects)
2. We already have code for this conversion that exists within the Python Code that could be used here or made a service for the front-end?: https://github.com/Zhao-Group/Primer_Design_and_Worklists/blob/hardcode_to_args/Primer_Design_SDM.py#L105
3. We already have code for switching between code translation tables within the Python code that could be made a front-end service?: https://github.com/Zhao-Group/Primer_Design_and_Worklists/blob/hardcode_to_args/Primer_Design_SDM.py#L79

[David-Bianchi]

Design (UI/UX) context:

[avireusdev11]

AC1 ✅
AC2 ✅
AC3 ✅
AC4 ✅
AC5 ✅

[David-Bianchi]

For AC5.
We can use from Bio.Data import CodonTable for getting the stop codons most likely or another biopython functionality.

[David-Bianchi]

For AC1.
For upstream-CDS-downstream we would append in this order.

We do this on the cDNA level:

As strings:

• upstream cDNA + CDS cDNA + downstream cDNA
• convert to amino acid (to show "translated sequence") - biopython package (design_primer() or an easier one)

[David-Bianchi]

For AC6.

Validate that:

• upstream/left overhang

• • is at least 50 long
• • is DNA

• downstream/right overhang

• • is at least 50 long
• • is DNA

From @Nilmani-Singh:

• For upstream and downstream, discard anything more than 50bp

• • first validate that minimum 50bp
• • only take the first 50bp to use in the algorithm

• default can be no upstream/downstream, OR, random up/downstream generated by us.

• • If user doesn't enter anything we not add any
• • instead of randomly generating cDNA of len 50.

[David-Bianchi]

@David-Bianchi

Add another AC in Issue: #4

to related to this point: https://uillinoisedu-my.sharepoint.com/:p:/r/personal/nsingh16_illinois_edu/_layouts/15/Doc.aspx?sourcedoc=%7B69146A71-68B5-47AB-8190-249901664535%7D&file=Site_Directed_Mutagenesis_NCSA.pptx&action=edit&mobileredirect=true

"But, if the mutation list has a mutation at amino acid positions <8 from start or >end-8, we must force the users to supply the up and down sequences, and random upstream/downstream will generate errors."

[avireusdev11]

For AC1, link to Biopython page - https://biopython.org/docs/1.75/api/Bio.Seq.html

[nsingh16-Illinois]

Here is what I think about the input design.

The best practice is consistency, So​.....

• For input, on frontend website show three boxes, left flank - cDNA - right​ flank
• Left and right sequences are optional, cDNA will be required​.
• There will be info in this area "for mutation close to start/end of cDNA, enter left/right sequences for correct primer design."​

This strategy should make things better.

[nsingh16-Illinois]

Additionally

We should add random 50bp DNA to cDNA if there is no left/right input.

If there’s no left input and a primer is requested at position 3, it would require accessing string positions like −10 or −15, which don’t exist. To prevent potential errors or glitches, we add a random 50 bp sequence. The user has already been warned that this may result in incorrect primers in such cases.