Skip to main content

Table 1 Parameters used in the different models

From: Integrative analysis of the Trypanosoma brucei gene expression cascade predicts differential regulation of mRNA processing and unusual control of ribosomal protein expression

  

BS-A

BS-B

BS-C

BS-D

BS-E

BS-F

PC-A

PC-B

a

No genes examined

851

851

642

851

851

851

3776

3776

b

μ (min−1) Division time

0.0019 (6 h)

0.0019 (6 h)

0.0019 (6 h)

0.0019 (6 h)

0.0019 (6 h)

0.0019 (6 h)

0.0011 (10.5 h)

0.0011 (10.5 h)

c

v (molecules.cell−1 .min−1)

0.24

0.24

0.24

0.357

0.278

0.278

0.24

0.234

d

k1

0.41

data

data

0.41

0.41

data

0.41

0.41

e

k2

0.41

0.41

data

0.41

0.41

0.41

0.41

0.41

f

Length adjustment α

z = 600

z = 600

z = 600

z = 660

§

§

z = 600

§

g

Correlation coefficient (r)

0.59

0.60

0.43

0.59

0.54

0.55

0.66

0.60

  1. The Table shows the parameters used for the various models. See also Fig. 1
  2. a) To optimise the model we chose genes for which we had reliable measurements of both half-life and mRNA abundance. For bloodstream forms, we had attempted to use RNASeq data to determine the rate at which the sequences immediately upstream of the splice site disappeared, with the hope of using that as a proxy for the splicing rate [16]. For this form we restricted the dataset to mRNAs that had measured precursor half-times of 1–5 min. For BS-C we narrowed this evern further, using only genes that also had a similar measurement for the gene immediately downstream. This means that the correlation coefficient for BS-C is not directly comparable to those for the other BS models
  3. (b) Based on measured growth rates (see also [17])
  4. (c) The transcription rate (v) is influenced by the initiation and elongation rates; since neither has been measured, the transcription rate was originally calculated for bloodstream forms to obtain the measured steady-state mRNA level for PGKC [17] and this value was used in models BS-A - BS-C. For models BS-D and BS-E the rate was adjusted to give a better fit for the dataset (BS-D) and for BS-E and BS-F, to give a better fit with the new length adjustment (f). For PC-A the rate was left the same as in BS-A. For PC-B we multiplied the adjusted rates from BS-E and BS-F by a factor of 0.84, based on the fact that the DNA replication rate in procyclic forms is 0.84 times that of bloodstream forms [56]
  5. (d) k1 = 0.41 is equivalent to a splicing half-time of 1.7 min. This is based on reliable individual measurements of splicing for 3 mRNAs, which gave estimates of 1–3 min [14, 17, 31]. The “data” are the rates estimated by RNASeq [16] (see (a))
  6. (e) Data for BS-C were from the gene immediately downstream (see (a))
  7. (f) For models with a z value: α = (mRNA_length)/z as used in [16]. For models marked §: α = 1/(20.603 x (mRNA_length-0.461) for bloodstream forms; α = 1/(343.77 x (mRNA_length-0.858) for procyclics. The change in the spliced intermediate over time is:
  8. d([5′ spliced intermediate]/dt) = k1 x [precursor] - (k2 + α k4 + μ) x [5′ spliced intermediate]
  9. (g) Pearson’s correlation coefficient (r) between the abundances predicted for individual RNAs, and the measured abundances of those mRNAs
  10. The mRNA degradation (k5) values for the mRNAs were those measured in [16]. The precursor degradation rates k3 and k4 were always 0.08 (half-life 8 min) which is the rate of disappearance of the PGKB precursor when splicing is inhibited by Sinefungin in bloodstream forms [17]
Back to article page

BMC Genomics

ISSN: 1471-2164

Contact us