H2MM analysis Matlab files

Matlab files including all variables and results of H2MM analysis. The variables are:
1. burstData - the imported burst properties table2. BGtraj - the important 2s trajectory of BG rates3. photonID - the imported photon stream identity file4. timestamps - the imported photon timestamps file (in 12.5 ns clock units)5. Data - a list if bursts. Each burst includes an array of burst photons with 2 columns, photon timestamp relative to beginning of burst and photon stream identity6. additional input variables per each state model (here as example, for 4 states):6a. obsmati_4s - the initial values (to be optimized) of the probability for donor (1st column) or acceptor (2nd column) photon per each state (rows). sum of two columns per a row has to be 1.6b. priori_4s - the initial values of the state fraction. sum has to be 1.6c. transmati_4s - the initial rate matrix in units of ns^-17. output variables of H2MM analysis per each state model (here as example, for 4 states):
7a. obsmat1_4s - the most likely values of the probability for donor (1st column) or acceptor (2nd column) photon per each state (rows). sum of two columns per a row has to be 1.7b. prior1_4s - the most likely values of the state fraction. sum has to be 1.7c. transmati_4s - the most likely rate matrix in units of ns^-17d - transmat1_4s_Seconds - the most likely rate matrix in units of s^-17e. LL1_1_4s - the last value of this array holds the maximal log-likelihood, L, at which H2MM analysis converged. this value is used for the calculation of the BIC (Eq. S9). Together with the minimal BIC (usually for a 4-state model), BIC' is calculated (Eq. S11).7f. converged1_1_4s - gains a value of 1 when H2MM analysis converges, otherwise gains a value of 08 - sm - includes the total number of photons in all bursts, n (see Eq. S9).
The output variables of the H2MM analysis, obsmat, prior and transmat, are used as input parameters in the Viterbi algorithm, to calculate the state posterior probabilities and the most likely state per each photon in all bursts. The output variables of the Viterbi algorithm per each state model (here as example, for 4 states) are:9. gamma_4s - a list of bursts. Each list item includes the posterior probability of each state (1 through 4; a row for a state) for each photon (column)10. viterbi_path_4s - a list of bursts. Each list item includes the most probable state for each photon (column)
The output variables of the Viterbi algorithms are used as input in dwell analysis. The output variables of dwell analysis per each state model (here as example, for 4 states) are:11. array_of_bursts_with_transitions_4s - an array that include the properties of dwells (a row) that can be considered for dwell time analysis:11a. The dwell is in within Burst ID11b. The ID number of the first photon in a dwell, relative to the photon IDs in a burst11c. The ID number of the last photon in a dwell, relative to the photon IDs in a burst
11d. The dwell time, in seconds, using 11b. and 11c.11e. The ID number of one photon before the first photon of the dwell11f. The ID number of one photon after the last photon of the dwell11g. The dwell time, in seconds, using 11e. and 11f.11h. The state to which the dwell, i, belongs11i. The state from which the dwell i-1 transited into the current dwell i11j. The state to which the dwell i transitions to, dwell i+1
12. dwell_information_4s - dwell information, for all dwells, including those that do not start or do not end in within a burst. The properties of dwells (a row) that can be considered for dwell time analysis:12a. Like 11a.12b. Like 11b.12c. Like 11c.12d. The dwell size - how many DexDem and DexAem photons it includes.12e. Dwell time in seconds12f. The proximity ratio (PR) value of the dwell, not corrected to BG12g. The state number to which the dwell belongs12h. The PR value of the dwell, corrected for BG (Eq. S3)12i. The corrected E value of the dwell (corrected for BG, lk, dir and gamma; Eq. S4)12j. Like 12d., but corrected for BG. Used as a weighting factor for the dwell, n'. See Eq. S12.
Dwell FRET statistics - variables derived from statistics over all dwells per each state in a given state model (here as example, for 4-states model, for the 3rd state):
13. FRETvalues_4s_state3 - values of PR uncorrected for BG (row 1), PR corrected for BG (row 2) and corrected E (row 3), for no weights (columns 1-2, column 2 holds the SE values) or after dwell-size weighting (columns 3-4, column 4 holds the SE values, Eqs. S12 and S13). The values in Table S2, in Fig. 4 and in the input for FPS analysis are derived from row 3, columns 3 & 4.
14. FREThist_4s_state3 - FRET histograms. columns:column 1 - PR/E binscolumn 2 - dwell PR histogram uncorrected and unweightedcolumn 3 - dwell PR histogram uncorrected dwell-size weightedcolumn 4 - dwell PR histogram BG corrected and unweightedcolumn 5 - dwell PR histogram BG corrected and dwell-size weightedcolumn 6 - dwell E histogram unweightedcolumn 7 - dwell E histogram dwell-size weighted. This is the column used for the presentation of dwell histograms in Fig. 3
Mean dwell times and rate constants - variables derived from statistics over all dwells per each state in a given state model (here as example, for 4-states model, for the transition between the 3rd and 2nd states; Eqs. S14 and S15):
15. dwells_4s_3to2_mean - holds the mean dwell time in state 3, before transitioning to state 2 (in seconds)16. dwells_4s_3to2_se - holds the SE of the dwell time in state 3, before transitioning to state 2 (in seconds)17. k32_mean - holds the mean rate constant for the 3->2 transition (in s-1)18. k32_se - holds the SE of the rate constants for the 3->2 transition (s-1)
Additional variables:
19. lk. holds the value of the lk factor20. dir. holds the value of the dir factor21. gammaFactor. holds the value of the Gamma factor found for the D-A construct22. FREThist - holds the values of the burst-derived FRET histogram, stemming from the burstData variable.