SiMD - Siriraj Metabolomics Data Warehouse

1. Overview

Siriraj Metabolomics Data Warehouse for Ion Mobility-Mass Spectrometry (SiMD) is a four-dimensional (4D) library that encompasses mass-to-charge ratio (m/z), retention time (RT), fragment ions (MS/MS), and collision cross-section (CCS) values, all derived from reference standards characterized by ion mobility-mass spectrometry. The SiMD library contains both human metabolites and specialized metabolites from natural products. This advanced mass spectral database provides scoring algorithms for different matching parameters to facilitate various levels of metabolite identification for both traveling wave ion mobility-mass spectrometry (TWIMS) and cyclic ion mobility-mass spectrometry (cIMS).

2. Search

SiMD supports metabolite identification for both traveling wave ion mobility-mass spectrometry (TWIMS) and cyclic ion mobility-mass spectrometry (cIMS). It allows both single and batch (CSV) searches. For each platform, SiMD offers 11 search schemes (see Tables 1 and 2).

Table 1. Search schemes for TWIMS

No.	Search schemes for TWIMS	Description
1	Name	Name searching e.g., Butyric acid
2	Xref	Cross reference searching e.g., HMDB0000039
3	m/z	m/z searching
4	CCS	CCS searching
5	m/z + MS/MS	m/z and MS/MS searching
6	m/z + CCS	m/z and CCS searching
7	m/z + RT	m/z and RT searching
8	m/z + CCS + RT	m/z, CCS and RT searching
9	m/z + CCS + MS/MS	m/z, CCS and MS/MS searching
10	m/z + RT + MS/MS	m/z, RT and MS/MS searching
11	m/z + CCS +RT + MS/MS	m/z, CCS, RT and MS/MS searching

Table 2. Search schemes for cIMS

No.	Search schemes for cIMS	Description
1	Name	Name searching e.g., Butyric acid
2	Xref	Cross reference searching e.g., HMDB0000039
3	m/z	m/z searching
4	CCS	CCS searching
5	m/z + MS/MS	m/z and MS/MS searching
6	m/z + CCS	m/z and CCS searching
7	m/z + t_p	m/z and t_p searching
8	m/z + t_n	m/z and t_n searching
9	m/z + CCS + MS/MS	m/z, CCS and MS/MS searching
10	m/z + t_p + MS/MS	m/z, t_p and MS/MS searching
11	m/z + t_n + MS/MS	m/z, t_n and MS/MS searching

3. Search Criteria

Users can define the adduct (e.g., [M+H]+ and [M-H]-) as search criteria. They can also set tolerance for m/z, CCS, RT, MS/MS, t_p and t_n searching. The following equations are used for setting tolerance for each parameter.

(1)Δm/z = Parts per million (ppm) measures the approximation error between the experimental mass and the reference mass

$Δm/z = \frac{| {mz}_{input} - {mz}_{ref} |}{{mz}_{ref}} \times 10^{6}$

(2)ΔCCS = Deviation of CCS (%)

$ΔCCS = \frac{| {CCS}_{input} - {CCS}_{ref} |}{{CCS}_{ref}} \times 100$

(3)ΔRT = Deviation of RT (min)

$ΔRT = | {RT}_{input} - {RT}_{ref} |$

(4)Δtₚ = Deviation of tₚ (%)

${Δt}_{p} = \frac{| t_{p (input)} - t_{p (ref)} |}{t_{p (ref)}} \times 100$

(5)Δtₙ = Deviation of tₙ (%)

${Δt}_{n} = \frac{| t_{n (input)} - t_{n (ref)} |}{t_{n (ref)}} \times 100$

4. Similarity Score

The following equations are used to calculate a scaled similarity score for m/z, CCS, RT, t_p, t_n. The similarity scores range between 0 and 100, with 100 signifying a perfect match. A scaling value (X) for each parameter can be specified by the user. The default values of X_mz = 2, X_CCS = 10, X_RT = 200, X_tp = 20, X_tn = 20 will result in a score of 0, if Δm/z = 50, ΔCCS = 10, Δtp = 5, Δtn = 5, respectively. A lower scaling value will result in a higher score. This helps determine the significance when searching using more than one parameter.

(6)m/z similarity score

mz_sim = 100 - (X_mz × Δm/z) ; X_mz = 2

(7)CCS similarity score

CCS_sim = 100 - (X_CCS × ΔCCS) ; X_CCS = 10

(8)RT similarity score

RT_sim = 100 - (X_RT × ΔRT) ; X_RT = 200

(9)t_p similarity score

${tp}_{sim} = 100 - (X_{tp} \times Δ tp); X_{tp} = 20$

(10)t_n similarity score

${tn}_{sim} = 100 - (X_{tn} \times Δ tn); X_{tn} = 20$

(11)Average similarity score when searching more than one parameter

$Average Similarity Score = \frac{\sum (similarity scores)}{N}; N - number of search parameters$

(12)MS/MS similarity score when peak intensity or peak height is not considered

$MS / MS Similarity Score1 = \frac{number of matched mz}{number of mz_{ref}} \times 100$

(13)MS/MS similarity score when peak intensity or peak height is considered

SiMD uses a weighting function (W) where the m/z value is raised to the power of m, and the peak intensity is raised to the power of n. The values of m and n can be specified by the user to determine the significance of m/z and peak intensity.

$W = {mz}^{m} \times {Intensity}^{n}; m = 1 and n = 0.6$
$MS / MS Similarity Score2 = \frac{\sum (W_{input} \times W_{ref})}{\sqrt{\sum (W_{input}^{2}) \times \sum (W_{ref}^{2})}}$