Craniofacial Research Capstone

Question: How conserved is the human IRF6 protein in mouse and zebrafish, and what does that say about its importance?

• Claim: IRF6 is highly conserved, which means it is doing an essential job.
• Evidence: My blastp of human IRF6 found the mouse protein at about 94 percent identity and the zebrafish protein at about 70 percent identity, both with E-values near zero.
• Reasoning: Humans and zebrafish last shared an ancestor over 400 million years ago. If the sequence is still about 70 percent identical after all that time, harmful mutations must have been removed by selection because the gene matters. A gene you can freely mutate would have drifted far more. So the conservation is evidence that IRF6 has an essential role, which fits its link to building the face.

Question: How would a pathogenic IRF6 variant in the DNA-binding domain change the protein's function?

• Setup: On UniProt, the DNA-binding domain of IRF6 sits near the start of the protein. AlphaFold models it as a compact, high-confidence fold of beta strands and helices that forms the surface that contacts DNA.
• Variant: I chose a pathogenic missense variant that swaps one amino acid for another inside that domain.
• Prediction: Because the changed residue sits in the part of the fold that grips DNA, the substitution likely distorts that surface. If the shape is wrong, IRF6 cannot bind its target DNA well, so it cannot switch its genes on at the right time. During the weeks when the lip and palate fuse, that lost regulation could leave the tissue without the signal it needs, which fits why this variant causes clefting.

Question: How are the reported IRF6 variants distributed, and what does the share of uncertain ones tell me?

• Data: I sorted the ClinVar IRF6 entries by classification and counted each group (numbers below are illustrative of the pattern I found).
• Pattern: A large block of variants is labeled Uncertain significance, alongside a clear set of Pathogenic ones.
• My question and partial answer: Why are so many uncertain? Most likely because they are rare and have not been seen in enough patients or studied in enough lab experiments to classify. That is a real research gap: each VUS is a candidate someone could investigate. A good next step would be to check whether the pathogenic variants cluster in the DNA-binding domain, which would predict where future VUS in that region are more likely to be harmful.

Question: Do the CDC data show a difference in how the types of orofacial clefts occur?

• Data: From the CDC facts page I recorded the reported prevalence of cleft lip with or without cleft palate and of cleft palate alone, and I noted the page's statement that cleft lip is more common in boys.
• Chart: I made a bar chart of the two cleft types side by side so the difference is easy to see.
• Claim: The two cleft types do not occur at the same rate, and the data also describe a sex difference for cleft lip.
• Limit: These are associations. The chart shows that the pattern exists, but it does not by itself prove why; testing a cause would need the kind of gene and developmental studies in the other missions.

Question: Which fusion windows must have been disrupted to produce Mateo's cleft?

• Model: My timeline shows lip fusion happening earlier (about weeks 4 to 7) and palate fusion later (about weeks 6 to 12).
• Mapping: A cleft lip maps to the earlier lip-fusion window; an isolated cleft palate maps to the later palate window.
• Prediction: Mateo has both a cleft lip that reaches into the nose and an opening in the palate. So my model predicts that the left-side lip prominences failed to fuse in the earlier window AND the palatal shelves failed to fuse in the later window. The lip event is on the left only, which fits a one-sided failure of the prominences on that side. This is testable: if I learned the exact timing was even earlier, I would expect the lip cleft to be more severe.

Question: Do pathogenic IRF6 variants cluster in the DNA-binding domain more than uncertain variants do?

• Why it is feasible: I can answer it with free tools (ClinVar for the variant list and classifications, UniProt for where the DNA-binding domain sits).
• Independent variable: classification of each variant (Pathogenic versus Uncertain significance).
• Dependent variable: whether the variant's position falls inside the DNA-binding domain.
• Control / comparison: the uncertain-significance variants act as the comparison group against the pathogenic ones, so I can see whether clustering is specific to the harmful ones.
• What I would conclude: If pathogenic variants land in the DNA-binding domain far more often than uncertain ones, that supports the idea that disrupting DNA binding is a main way IRF6 causes clefting, and it predicts that future uncertain variants in that domain deserve a closer look.
• Limit: ClinVar classifications can change as evidence grows, and my sample is only the variants reported so far.