Out of the 27 genes that we shortlisted, we were able to obtain products from a PCR off extracted M. polymorpha DNA with a success rates of 50% - 70% on successive runs. We ran out of time to satisfactorily complete this stage and debug our difficulties with Gibson Assembly, but we think that we have good evidence that these candidates warrant further investigation.

Notes:
• In Figures 2 and 3 the predicted gene numbers are grouped according to the related essential nutrient: genes 603 – 11107 inclusive relate to nitrates; genes 3555 and 3556 relate to sulphates and 170-11169 inclusive relate to phosphates.
• The percentage overlaps in Figure 3 measure how much of a gene sequence length was covered by the hit sequence length, and how much of the hit sequence length was covered by an mRNA sequence length in a 1-to-1 mapping between nucleotides.

The mean total number of hits for a predicted gene was 5 [see Figure 2]. Of the genes for which the hit value is at this frequency or above, all except three relate to nitrate-inducible responses; the remainder relate to potassium starvation responses. It can be considered that the queried genes with more matches to M. polymorpha along the predicted gene sequence lengths are likelier to be better homologues. However we think that a mapping of the blast® hits to the predicted genes would be needed to confirm this. This is because a spread of hits along the gene length would give stronger evidence than many hits to the same location, which we have not ruled out at this time.

Comparing Figures 2 and 3 , predicted genes 4354 and 11107 are the best candidates for further work as they both have relatively high hit frequency and congruence with mRNA and the queried gene sequences. Gene 11107 also demonstrates a good match with the gene AtNRT2.1 (45.6% congruence) which has been shown to be “induced by very low levels of nitrate (50 μM KNO3)” (Filleur, Daniel-Vedele, 1999)[8]. Given this sensitivity, we suggest that this gene would be a very good candidate for experimentally verifying the associated promoter (P_Ni20, iGEM Registry Part:BBa_K1484336). Gene 4354 demonstrates a good match with the gene AtNRT2.2 (57.7% congruence), which is expressed up to 8 days earlier than AtNRT2.1 (Zhuo et al., 1999)[9] and so it may be quicker to experimentally verify the promoter for this potential homologue [P_Ni17, iGEM Registry Part:BBa_K1484333].

Although gene 11169 did not match with an mRNA transcript, it has been included in Figure 3 as an observation of interest. It might seem a very good candidate for further work given the startling degree of congruence between the hit and the gene, but we were yet to verify the reliability of the reconstructed regions of the genome scaffolds where mRNA sequence data was unavailable.

We also observed that the majority of the better hits fell along middling regions of the predicted genes. This seems reasonable given that Arabidopsis thaliana emerged approximately 10 million years ago[10] and Marchantia emerged around 400 million years ago[1], so there would have been sufficient time for non-essential amino acids to get knocked out or replaced by genetic drift, with the essential ones (e.g. molecule binding motifs) being conserved.

Conclusion